WEBVTT 00:00:00.000 --> 00:00:01.948 align:middle line:90% [SQUEAKING] 00:00:01.948 --> 00:00:03.896 align:middle line:90% [RUSTLING] 00:00:03.896 --> 00:00:07.305 align:middle line:90% [CLICKING] 00:00:07.305 --> 00:00:09.452 align:middle line:90% 00:00:09.452 --> 00:00:11.660 align:middle line:84% MICHAEL CUTHBERT: Today, I'm going to be a little bit 00:00:11.660 --> 00:00:14.240 align:middle line:84% boring for people watching later and stuff 00:00:14.240 --> 00:00:17.250 align:middle line:84% because we're going to mostly be flipping things around. 00:00:17.250 --> 00:00:19.400 align:middle line:84% So we're going to use this time for quite 00:00:19.400 --> 00:00:21.985 align:middle line:90% a lot of in-class work. 00:00:21.985 --> 00:00:25.070 align:middle line:90% Hey, how you doing? 00:00:25.070 --> 00:00:26.300 align:middle line:90% You might need a chair there. 00:00:26.300 --> 00:00:29.570 align:middle line:90% And try to get things-- 00:00:29.570 --> 00:00:32.270 align:middle line:84% getting ready for later, and having videos 00:00:32.270 --> 00:00:38.750 align:middle line:84% online for some of the new topics that are coming up. 00:00:38.750 --> 00:00:40.760 align:middle line:90% Before we get to-- 00:00:40.760 --> 00:00:43.490 align:middle line:84% this is the fifth week of the semester. 00:00:43.490 --> 00:00:51.050 align:middle line:84% First, are there any questions so far on problem set four? 00:00:51.050 --> 00:00:51.780 align:middle line:90% Adam? 00:00:51.780 --> 00:00:53.947 align:middle line:84% AUDIENCE: So with OPTIC, you would want the two list 00:00:53.947 --> 00:00:56.930 align:middle line:90% at the same length? 00:00:56.930 --> 00:00:58.170 align:middle line:90% MICHAEL CUTHBERT: Let me see. 00:00:58.170 --> 00:01:04.900 align:middle line:84% Would I want-- no, not necessarily. 00:01:04.900 --> 00:01:06.960 align:middle line:84% Let's think through, again, OPTIC. 00:01:06.960 --> 00:01:08.322 align:middle line:90% O is-- just shout it out. 00:01:08.322 --> 00:01:09.030 align:middle line:90% AUDIENCE: Octave. 00:01:09.030 --> 00:01:10.072 align:middle line:90% MICHAEL CUTHBERT: Octave. 00:01:10.072 --> 00:01:10.761 align:middle line:90% P? 00:01:10.761 --> 00:01:12.084 align:middle line:90% AUDIENCE: Permutation. 00:01:12.084 --> 00:01:12.966 align:middle line:90% MICHAEL CUTHBERT: T? 00:01:12.966 --> 00:01:13.410 align:middle line:90% AUDIENCE: Transposition. 00:01:13.410 --> 00:01:14.730 align:middle line:84% MICHAEL CUTHBERT: Transposition, good. 00:01:14.730 --> 00:01:15.247 align:middle line:90% I? 00:01:15.247 --> 00:01:16.080 align:middle line:90% AUDIENCE: Inversion. 00:01:16.080 --> 00:01:16.710 align:middle line:90% MICHAEL CUTHBERT: Inversion. 00:01:16.710 --> 00:01:17.580 align:middle line:90% And C? 00:01:17.580 --> 00:01:17.970 align:middle line:90% AUDIENCE: Cardinal. 00:01:17.970 --> 00:01:19.095 align:middle line:90% MICHAEL CUTHBERT: Cardinal. 00:01:19.095 --> 00:01:21.783 align:middle line:84% One of those is very important that you test-- 00:01:21.783 --> 00:01:23.700 align:middle line:84% with all of them, it's important that you test 00:01:23.700 --> 00:01:25.270 align:middle line:90% with different length sets. 00:01:25.270 --> 00:01:31.230 align:middle line:84% But what's one of the ones that might sometimes return true 00:01:31.230 --> 00:01:37.110 align:middle line:84% if the two pitch sets are equivalent if you 00:01:37.110 --> 00:01:40.483 align:middle line:90% have this equivalence class? 00:01:40.483 --> 00:01:41.400 align:middle line:90% AUDIENCE: Cardinality. 00:01:41.400 --> 00:01:42.900 align:middle line:84% MICHAEL CUTHBERT: Cardinality, yeah. 00:01:42.900 --> 00:01:47.220 align:middle line:84% So cardinality equivalence, it's kind of an odd thing, isn't it? 00:01:47.220 --> 00:01:50.830 align:middle line:84% Because it's not-- it's kind of the backwards one, isn't it? 00:01:50.830 --> 00:01:54.960 align:middle line:84% It should be like optic duplication or something 00:01:54.960 --> 00:01:57.430 align:middle line:84% because it is specifically not saying 00:01:57.430 --> 00:02:02.270 align:middle line:84% that things are equivalent if their cardinality is the same. 00:02:02.270 --> 00:02:06.435 align:middle line:84% It is saying two things are equivalent if they're-- 00:02:06.435 --> 00:02:07.310 align:middle line:90% what do you want to-- 00:02:07.310 --> 00:02:12.280 align:middle line:84% I suppose cardinality is that if duplicates are removed. 00:02:12.280 --> 00:02:12.820 align:middle line:90% Good. 00:02:12.820 --> 00:02:16.270 align:middle line:84% So yeah, there will be cases where 00:02:16.270 --> 00:02:21.970 align:middle line:84% two pitch sets will have different lengths of things 00:02:21.970 --> 00:02:25.420 align:middle line:90% and be equivalent. 00:02:25.420 --> 00:02:28.900 align:middle line:90% Oh, one important thing. 00:02:28.900 --> 00:02:32.170 align:middle line:84% I'll do the one that I forgot to do. 00:02:32.170 --> 00:02:34.690 align:middle line:90% That's on inversion. 00:02:34.690 --> 00:02:41.560 align:middle line:84% An inversion is one of these things-- 00:02:41.560 --> 00:02:44.200 align:middle line:84% probably like mode or something like that-- 00:02:44.200 --> 00:02:48.370 align:middle line:84% we used way too often in music to mean too many 00:02:48.370 --> 00:02:50.460 align:middle line:90% related but different things. 00:02:50.460 --> 00:02:54.370 align:middle line:84% I'm just going to move that more to the center. 00:02:54.370 --> 00:03:00.320 align:middle line:84% So the two places that we've been talking about inversion 00:03:00.320 --> 00:03:05.330 align:middle line:90% are in this-- 00:03:05.330 --> 00:03:10.520 align:middle line:84% do you remember the circle from C, C sharp slash D flat. 00:03:10.520 --> 00:03:11.630 align:middle line:90% Spelling doesn't matter. 00:03:11.630 --> 00:03:13.980 align:middle line:84% D-- am I getting there about right? 00:03:13.980 --> 00:03:14.480 align:middle line:90% Yep. 00:03:14.480 --> 00:03:18.390 align:middle line:90% E flat, E, F. Now somehow I've-- 00:03:18.390 --> 00:03:22.310 align:middle line:84% well, because my circle is more of an egg. 00:03:22.310 --> 00:03:27.993 align:middle line:84% So on, and-- sorry, I'm blanking on-- 00:03:27.993 --> 00:03:29.660 align:middle line:84% it's been a long time since I've written 00:03:29.660 --> 00:03:32.660 align:middle line:84% a circle of notes that is not the circle of fifths 00:03:32.660 --> 00:03:33.720 align:middle line:90% on the board. 00:03:33.720 --> 00:03:35.790 align:middle line:90% OK, got that in order. 00:03:35.790 --> 00:03:39.380 align:middle line:84% And we've thought about this axis of inversion 00:03:39.380 --> 00:03:40.830 align:middle line:90% that you can have. 00:03:40.830 --> 00:03:43.040 align:middle line:90% Do you remember that? 00:03:43.040 --> 00:03:46.130 align:middle line:84% That can define ways that pitches 00:03:46.130 --> 00:03:47.940 align:middle line:90% can invert to each other. 00:03:47.940 --> 00:03:50.760 align:middle line:90% We also use inversion. 00:03:50.760 --> 00:03:54.520 align:middle line:84% And so this is the primary one that Dimitri Tymoczko 00:03:54.520 --> 00:03:55.580 align:middle line:90% is talking about. 00:03:55.580 --> 00:03:57.940 align:middle line:84% We also use inversion to mean something which 00:03:57.940 --> 00:04:00.080 align:middle line:90% is called tonal inversion. 00:04:00.080 --> 00:04:02.110 align:middle line:90% Which is to say-- 00:04:02.110 --> 00:04:02.690 align:middle line:90% oh, yeah. 00:04:02.690 --> 00:04:04.480 align:middle line:84% We did this in class for a second, right? 00:04:04.480 --> 00:04:12.820 align:middle line:84% That this can be inverted to be this. 00:04:12.820 --> 00:04:15.700 align:middle line:90% 00:04:15.700 --> 00:04:19.240 align:middle line:84% Even though-- so because here, we go up a step. 00:04:19.240 --> 00:04:20.240 align:middle line:90% Here, we go down a step. 00:04:20.240 --> 00:04:21.380 align:middle line:90% Here we go up another step. 00:04:21.380 --> 00:04:24.490 align:middle line:84% We go down a step, up a step, down a step. 00:04:24.490 --> 00:04:27.050 align:middle line:84% Even though the steps are different sizes, right? 00:04:27.050 --> 00:04:29.500 align:middle line:90% So F to G is what kind of step? 00:04:29.500 --> 00:04:30.670 align:middle line:90% Whole step. 00:04:30.670 --> 00:04:31.720 align:middle line:90% Next one. 00:04:31.720 --> 00:04:32.475 align:middle line:90% Whole step. 00:04:32.475 --> 00:04:33.100 align:middle line:90% AUDIENCE: Half. 00:04:33.100 --> 00:04:34.058 align:middle line:90% MICHAEL CUTHBERT: Half. 00:04:34.058 --> 00:04:34.805 align:middle line:90% And this one? 00:04:34.805 --> 00:04:35.430 align:middle line:90% AUDIENCE: Half. 00:04:35.430 --> 00:04:37.460 align:middle line:84% MICHAEL CUTHBERT: Half, whole, whole. 00:04:37.460 --> 00:04:44.600 align:middle line:84% But so in the strictly enharmonic context of inversion, 00:04:44.600 --> 00:04:46.390 align:middle line:84% they would not be the same, which 00:04:46.390 --> 00:04:48.970 align:middle line:84% is the context that Dimitri Tymoczko is mostly 00:04:48.970 --> 00:04:50.035 align:middle line:90% talking about. 00:04:50.035 --> 00:04:55.240 align:middle line:84% In the tonal context, they would be the same though, right? 00:04:55.240 --> 00:04:59.340 align:middle line:84% But then there's this third part, this third very common use 00:04:59.340 --> 00:05:02.910 align:middle line:84% of the word inversion in music that 00:05:02.910 --> 00:05:07.080 align:middle line:84% is not what we're talking about in any way on this problem set. 00:05:07.080 --> 00:05:08.715 align:middle line:90% John, do you know what that is? 00:05:08.715 --> 00:05:09.840 align:middle line:90% AUDIENCE: Chords inversion. 00:05:09.840 --> 00:05:11.257 align:middle line:84% MICHAEL CUTHBERT: Chord inversion. 00:05:11.257 --> 00:05:12.600 align:middle line:90% Chord inversion. 00:05:12.600 --> 00:05:13.210 align:middle line:90% Right. 00:05:13.210 --> 00:05:15.840 align:middle line:90% So if I have a chord-- 00:05:15.840 --> 00:05:17.503 align:middle line:90% let's say, A major-- 00:05:17.503 --> 00:05:21.090 align:middle line:90% 00:05:21.090 --> 00:05:26.630 align:middle line:84% what is the base if I am in second inversion? 00:05:26.630 --> 00:05:27.570 align:middle line:90% AUDIENCE: E? 00:05:27.570 --> 00:05:30.810 align:middle line:90% MICHAEL CUTHBERT: E, great. 00:05:30.810 --> 00:05:33.540 align:middle line:84% So you might have something like that. 00:05:33.540 --> 00:05:36.380 align:middle line:84% Now here-- so that's the explanation part. 00:05:36.380 --> 00:05:37.630 align:middle line:90% That's the clarification part. 00:05:37.630 --> 00:05:39.900 align:middle line:84% And now the critical thinking thing. 00:05:39.900 --> 00:05:43.380 align:middle line:84% Chord inversion, in this sense, has 00:05:43.380 --> 00:05:46.410 align:middle line:90% to do with what letter of optic? 00:05:46.410 --> 00:05:48.960 align:middle line:90% 00:05:48.960 --> 00:05:51.715 align:middle line:90% AUDIENCE: P? 00:05:51.715 --> 00:05:54.410 align:middle line:84% MICHAEL CUTHBERT: P. And there's another one. 00:05:54.410 --> 00:05:58.720 align:middle line:84% So P-- we could say that A, C sharp, E is 00:05:58.720 --> 00:06:04.190 align:middle line:84% the same as E, C sharp, A, but there's something else we need. 00:06:04.190 --> 00:06:05.060 align:middle line:90% Yeah? 00:06:05.060 --> 00:06:05.560 align:middle line:90% AUDIENCE: O? 00:06:05.560 --> 00:06:08.920 align:middle line:84% MICHAEL CUTHBERT: O. So we need that this O 00:06:08.920 --> 00:06:12.490 align:middle line:84% is equivalent to this O, that this octave is 00:06:12.490 --> 00:06:13.610 align:middle line:90% the same as this octave. 00:06:13.610 --> 00:06:14.902 align:middle line:90% I see everybody shivering. 00:06:14.902 --> 00:06:16.610 align:middle line:84% It's really cold in here today, isn't it? 00:06:16.610 --> 00:06:17.490 align:middle line:90% All right? 00:06:17.490 --> 00:06:17.990 align:middle line:90% OK. 00:06:17.990 --> 00:06:19.480 align:middle line:90% No, maybe it's just me. 00:06:19.480 --> 00:06:20.300 align:middle line:90% You were cold. 00:06:20.300 --> 00:06:22.010 align:middle line:84% Other people were cold, and I'm up here. 00:06:22.010 --> 00:06:22.760 align:middle line:90% OK. 00:06:22.760 --> 00:06:27.706 align:middle line:90% So O and octave and permutation. 00:06:27.706 --> 00:06:30.400 align:middle line:90% 00:06:30.400 --> 00:06:34.550 align:middle line:84% If you have an augmented triad, there's an interesting-- here, 00:06:34.550 --> 00:06:35.810 align:middle line:90% let's just keep playing. 00:06:35.810 --> 00:06:37.720 align:middle line:84% We'll make a simpler augmented triad 00:06:37.720 --> 00:06:41.170 align:middle line:84% that you might have thought of before. 00:06:41.170 --> 00:06:45.460 align:middle line:84% Then the different inversions have to do with-- 00:06:45.460 --> 00:06:47.660 align:middle line:84% you can think of them as O and P. 00:06:47.660 --> 00:06:49.890 align:middle line:84% But specifically for the augmented triad, 00:06:49.890 --> 00:06:53.035 align:middle line:84% there's another letter that comes into play. 00:06:53.035 --> 00:07:00.260 align:middle line:90% 00:07:00.260 --> 00:07:03.100 align:middle line:84% Let's get the sound of the augmented triad into our ears. 00:07:03.100 --> 00:07:06.306 align:middle line:90% [PIANO NOTES PLAYING] 00:07:06.306 --> 00:07:08.150 align:middle line:90% 00:07:08.150 --> 00:07:10.760 align:middle line:84% What's a way that we can do, let's say, 00:07:10.760 --> 00:07:18.380 align:middle line:84% a first inversion of this augmented triad? 00:07:18.380 --> 00:07:19.465 align:middle line:90% Yeah? 00:07:19.465 --> 00:07:21.090 align:middle line:84% Actually, I'm just calling on somebody. 00:07:21.090 --> 00:07:22.520 align:middle line:90% Yeah, go ahead. 00:07:22.520 --> 00:07:24.260 align:middle line:90% AUDIENCE: I? 00:07:24.260 --> 00:07:25.340 align:middle line:90% MICHAEL CUTHBERT: I? 00:07:25.340 --> 00:07:26.190 align:middle line:90% Let's see. 00:07:26.190 --> 00:07:30.300 align:middle line:84% Yes, because we can invert around this point. 00:07:30.300 --> 00:07:33.620 align:middle line:84% And what are the two intervals from here to here? 00:07:33.620 --> 00:07:35.180 align:middle line:90% AUDIENCE: The same. 00:07:35.180 --> 00:07:37.460 align:middle line:84% MICHAEL CUTHBERT: They're the same. 00:07:37.460 --> 00:07:41.264 align:middle line:90% Yes, but what are they? 00:07:41.264 --> 00:07:42.363 align:middle line:90% Yeah, go ahead. 00:07:42.363 --> 00:07:43.280 align:middle line:90% AUDIENCE: Major third? 00:07:43.280 --> 00:07:45.484 align:middle line:84% MICHAEL CUTHBERT: Major third and? 00:07:45.484 --> 00:07:46.440 align:middle line:90% AUDIENCE: Major third. 00:07:46.440 --> 00:07:47.310 align:middle line:84% MICHAEL CUTHBERT: And major third because you 00:07:47.310 --> 00:07:48.100 align:middle line:90% already said the same. 00:07:48.100 --> 00:07:48.420 align:middle line:90% Good. 00:07:48.420 --> 00:07:49.060 align:middle line:90% Sorry. 00:07:49.060 --> 00:07:52.560 align:middle line:84% Sometimes I forget that it was prefaced with something 00:07:52.560 --> 00:07:58.590 align:middle line:84% that required me to use the C operator to map major third 00:07:58.590 --> 00:07:59.830 align:middle line:90% to major third, major third. 00:07:59.830 --> 00:08:00.145 align:middle line:90% Right? 00:08:00.145 --> 00:08:00.645 align:middle line:90% Good. 00:08:00.645 --> 00:08:01.613 align:middle line:90% So it can be inversion. 00:08:01.613 --> 00:08:03.280 align:middle line:84% There's another way you can think of it. 00:08:03.280 --> 00:08:04.240 align:middle line:90% Yeah, Jason? 00:08:04.240 --> 00:08:04.590 align:middle line:90% AUDIENCE: Transpositon? 00:08:04.590 --> 00:08:05.923 align:middle line:90% MICHAEL CUTHBERT: Transposition. 00:08:05.923 --> 00:08:08.400 align:middle line:84% So if we go up, what do we have to go up? 00:08:08.400 --> 00:08:09.490 align:middle line:90% A major third. 00:08:09.490 --> 00:08:20.500 align:middle line:84% So C goes up to E. E goes up a major third to G sharp. 00:08:20.500 --> 00:08:25.410 align:middle line:84% And then G sharp goes up a major third to? 00:08:25.410 --> 00:08:26.820 align:middle line:90% AUDIENCE: C? 00:08:26.820 --> 00:08:32.130 align:middle line:84% MICHAEL CUTHBERT: First, it goes up a major third to B sharp. 00:08:32.130 --> 00:08:35.740 align:middle line:84% And then what other letter are we thinking of? 00:08:35.740 --> 00:08:36.240 align:middle line:90% AUDIENCE: S. 00:08:36.240 --> 00:08:37.870 align:middle line:90% MICHAEL CUTHBERT: S, spelling. 00:08:37.870 --> 00:08:41.559 align:middle line:84% So that the-- what do you call it? 00:08:41.559 --> 00:08:44.800 align:middle line:90% The B sharp goes to C natural. 00:08:44.800 --> 00:08:48.430 align:middle line:84% You can think of all of these different letters 00:08:48.430 --> 00:08:50.050 align:middle line:84% and different equivalence classes 00:08:50.050 --> 00:08:53.320 align:middle line:84% as quite often, depending on your context, different ways 00:08:53.320 --> 00:08:55.420 align:middle line:90% of getting to the same answer. 00:08:55.420 --> 00:08:57.230 align:middle line:84% The hard part of the problem set, 00:08:57.230 --> 00:08:58.870 align:middle line:84% though, is sometimes figuring out 00:08:58.870 --> 00:09:05.720 align:middle line:84% if you're given multiple of these letters together, 00:09:05.720 --> 00:09:10.570 align:middle line:84% you're saying that you have the O and P set. 00:09:10.570 --> 00:09:13.940 align:middle line:90% Can you apply O and then P? 00:09:13.940 --> 00:09:18.040 align:middle line:84% And is that the same as applying P and then O? 00:09:18.040 --> 00:09:19.930 align:middle line:84% Sometimes it will be, sometimes it-- 00:09:19.930 --> 00:09:23.870 align:middle line:84% or some of these combinations of letters, the order won't matter. 00:09:23.870 --> 00:09:26.020 align:middle line:84% And some of the combinations of letters 00:09:26.020 --> 00:09:31.660 align:middle line:84% will open up new possibilities that chaining one and then 00:09:31.660 --> 00:09:35.920 align:middle line:84% the other will not be sufficient for. 00:09:35.920 --> 00:09:40.995 align:middle line:84% Anybody far enough on the P set to encounter any of those? 00:09:40.995 --> 00:09:43.885 align:middle line:90% 00:09:43.885 --> 00:09:45.852 align:middle line:90% No? 00:09:45.852 --> 00:09:46.560 align:middle line:90% Some of you have. 00:09:46.560 --> 00:09:49.370 align:middle line:90% Some people have turned them in. 00:09:49.370 --> 00:09:54.230 align:middle line:84% Do you remember any of the ones that were like that? 00:09:54.230 --> 00:09:56.750 align:middle line:84% AUDIENCE: It was, like, three days ago. 00:09:56.750 --> 00:09:57.510 align:middle line:90% Yeah. 00:09:57.510 --> 00:10:02.673 align:middle line:84% I think octave and permutation didn't matter, I would say. 00:10:02.673 --> 00:10:04.340 align:middle line:84% MICHAEL CUTHBERT: Octave and permutation 00:10:04.340 --> 00:10:07.345 align:middle line:90% didn't matter which one you did. 00:10:07.345 --> 00:10:10.208 align:middle line:90% 00:10:10.208 --> 00:10:11.430 align:middle line:90% I think that's correct. 00:10:11.430 --> 00:10:11.930 align:middle line:90% Yeah. 00:10:11.930 --> 00:10:12.630 align:middle line:90% Yep. 00:10:12.630 --> 00:10:16.417 align:middle line:84% So I should be thinking there's how many combinations of-- 00:10:16.417 --> 00:10:17.250 align:middle line:90% what do you call it? 00:10:17.250 --> 00:10:19.590 align:middle line:90% There's five choose two. 00:10:19.590 --> 00:10:22.410 align:middle line:84% There's five factorial combinations. 00:10:22.410 --> 00:10:27.420 align:middle line:84% But we're just-- no, two to the fifth combinations with optic. 00:10:27.420 --> 00:10:31.020 align:middle line:84% But a lot of them are never used. 00:10:31.020 --> 00:10:36.920 align:middle line:84% You would never use, I don't know, cardinal. 00:10:36.920 --> 00:10:38.960 align:middle line:84% What's something you wouldn't use? 00:10:38.960 --> 00:10:41.700 align:middle line:84% Inversion without permutation quite often, 00:10:41.700 --> 00:10:42.880 align:middle line:90% or something like that. 00:10:42.880 --> 00:10:43.380 align:middle line:90% Yeah. 00:10:43.380 --> 00:10:45.000 align:middle line:90% Great. 00:10:45.000 --> 00:10:48.300 align:middle line:84% Or inversion and permutation without octave equivalents 00:10:48.300 --> 00:10:49.000 align:middle line:90% or something. 00:10:49.000 --> 00:10:49.980 align:middle line:90% Fantastic. 00:10:49.980 --> 00:10:53.590 align:middle line:84% I want to introduce you to another friend. 00:10:53.590 --> 00:10:55.080 align:middle line:84% We're going to have more as class 00:10:55.080 --> 00:10:58.210 align:middle line:84% goes on, these little one to two minute guest speakers. 00:10:58.210 --> 00:11:03.090 align:middle line:84% And here is somebody who went through this class like you. 00:11:03.090 --> 00:11:05.640 align:middle line:90% Let's see. 00:11:05.640 --> 00:11:08.550 align:middle line:84% SOPHIA SUN: Hi, I'm Sophia, and I worked with Mike 00:11:08.550 --> 00:11:12.880 align:middle line:90% with Music21 in 2016, 2017. 00:11:12.880 --> 00:11:17.070 align:middle line:84% We wrote this paper called Emotion Planning-- 00:11:17.070 --> 00:11:19.440 align:middle line:84% called Emotion Painting: Lyric Affect 00:11:19.440 --> 00:11:22.750 align:middle line:84% Musical Relationship in a Large Scale Lead-Sheet Corpus. 00:11:22.750 --> 00:11:25.170 align:middle line:84% And then the idea is there's this concept called 00:11:25.170 --> 00:11:28.620 align:middle line:84% word painting in classical music, where the music reflects 00:11:28.620 --> 00:11:29.920 align:middle line:90% what the lyric says. 00:11:29.920 --> 00:11:32.730 align:middle line:84% So rise, you will sing like rise. 00:11:32.730 --> 00:11:34.800 align:middle line:90% Or fall, you would sing, fall. 00:11:34.800 --> 00:11:37.260 align:middle line:84% And then we want to see how prevalent that is 00:11:37.260 --> 00:11:40.160 align:middle line:90% in, say, folk or popular music. 00:11:40.160 --> 00:11:44.920 align:middle line:84% We had this corpus of around 1,900 songs of folk music 00:11:44.920 --> 00:11:49.600 align:middle line:84% or popular music, and then we're able to analyze it with Music21. 00:11:49.600 --> 00:11:52.210 align:middle line:84% What we did is we used the Natural Language Toolkit 00:11:52.210 --> 00:11:55.490 align:middle line:84% and analyzed the emotion conveyed by the lyrics, 00:11:55.490 --> 00:11:57.880 align:middle line:84% and then looked at the corresponding notes 00:11:57.880 --> 00:11:59.470 align:middle line:90% in the music. 00:11:59.470 --> 00:12:02.000 align:middle line:84% We found a bunch of interesting things, 00:12:02.000 --> 00:12:06.610 align:middle line:84% like more emotive words that convey an emotion are more 00:12:06.610 --> 00:12:10.360 align:middle line:84% likely to fall on strong beat strengths-- 00:12:10.360 --> 00:12:13.580 align:middle line:84% like the first beat or the third beat on beat-- 00:12:13.580 --> 00:12:18.550 align:middle line:84% whereas words that doesn't convey emotion-- like is, and-- 00:12:18.550 --> 00:12:20.720 align:middle line:84% will fall on weaker beat strengths. 00:12:20.720 --> 00:12:23.260 align:middle line:84% And also interesting things like words 00:12:23.260 --> 00:12:27.370 align:middle line:84% with strong negative emotions like anger, disgust, or fear 00:12:27.370 --> 00:12:30.340 align:middle line:90% tend to fall on shorter notes. 00:12:30.340 --> 00:12:31.470 align:middle line:90% And that's what we did. 00:12:31.470 --> 00:12:33.977 align:middle line:90% 00:12:33.977 --> 00:12:35.560 align:middle line:84% MICHAEL CUTHBERT: So we're going to be 00:12:35.560 --> 00:12:40.750 align:middle line:84% moving on to the concept of corpus and corpora 00:12:40.750 --> 00:12:41.690 align:middle line:90% just a little bit. 00:12:41.690 --> 00:12:44.180 align:middle line:84% And so what do we do-- and statistics 00:12:44.180 --> 00:12:47.710 align:middle line:84% because the only reason to look at a huge number of pieces 00:12:47.710 --> 00:12:52.810 align:middle line:84% is to come up with some kind of statistical summary 00:12:52.810 --> 00:12:56.930 align:middle line:84% or extraction, or to find things that are outliers within it. 00:12:56.930 --> 00:12:58.880 align:middle line:84% And even if you're thinking-- oh, yeah, 00:12:58.880 --> 00:13:00.340 align:middle line:90% or to do deep learning stuff. 00:13:00.340 --> 00:13:03.130 align:middle line:84% And those who have taken artificial intelligence 00:13:03.130 --> 00:13:07.130 align:middle line:84% and deep learning classes, you'll know that at its heart, 00:13:07.130 --> 00:13:08.810 align:middle line:84% these are all just statistics, right? 00:13:08.810 --> 00:13:10.870 align:middle line:84% This is all just a whole bunch of numbers 00:13:10.870 --> 00:13:14.740 align:middle line:84% that are being trained and extracted from things. 00:13:14.740 --> 00:13:18.890 align:middle line:84% The corpus I was hoping for us to work with today, 00:13:18.890 --> 00:13:22.150 align:middle line:84% I haven't been able to get it working well enough 00:13:22.150 --> 00:13:24.980 align:middle line:90% to not crash on my system. 00:13:24.980 --> 00:13:29.380 align:middle line:84% And that's my bad for not having this up. 00:13:29.380 --> 00:13:32.740 align:middle line:84% But I want to point out to you, we're going to be talking about 00:13:32.740 --> 00:13:42.380 align:middle line:84% more-- it's a collection of, I think, 2,000 encoded lieds-- 00:13:42.380 --> 00:13:46.250 align:middle line:84% art songs, songs that are sung in concerts and things 00:13:46.250 --> 00:13:54.470 align:middle line:84% primarily by women and Black and other minority composers. 00:13:54.470 --> 00:13:56.970 align:middle line:84% You'll see that there are some people in here-- 00:13:56.970 --> 00:14:00.660 align:middle line:84% Beethoven-- who don't fit in any of those categories. 00:14:00.660 --> 00:14:05.930 align:middle line:84% And that was in order to have a comparison set in there. 00:14:05.930 --> 00:14:07.820 align:middle line:84% I don't know, I'll have to ask Mark 00:14:07.820 --> 00:14:10.260 align:middle line:84% Gotham, the great researcher who is doing this. 00:14:10.260 --> 00:14:12.920 align:middle line:84% I think also, he had really hoped to get 2,000 and had 00:14:12.920 --> 00:14:15.637 align:middle line:84% a certain day and found that he already had some Beethoven lying 00:14:15.637 --> 00:14:16.470 align:middle line:90% around or something. 00:14:16.470 --> 00:14:19.230 align:middle line:84% But it's a really amazing corpus, 00:14:19.230 --> 00:14:21.830 align:middle line:84% and I hope that we're going to be able to work with it later 00:14:21.830 --> 00:14:23.690 align:middle line:90% in the year. 00:14:23.690 --> 00:14:25.440 align:middle line:84% Today being International Women's Day, 00:14:25.440 --> 00:14:27.710 align:middle line:84% I thought that would be a particularly good day 00:14:27.710 --> 00:14:28.800 align:middle line:90% to work with it. 00:14:28.800 --> 00:14:33.030 align:middle line:84% But alas, things don't often hit our deadlines, 00:14:33.030 --> 00:14:37.085 align:middle line:84% and we can come back to that later. 00:14:37.085 --> 00:14:40.410 align:middle line:90% 00:14:40.410 --> 00:14:43.390 align:middle line:84% The second one that we're going to be working with-- 00:14:43.390 --> 00:14:46.050 align:middle line:84% and the reason why you have your laptops out today that 00:14:46.050 --> 00:14:48.250 align:middle line:84% we're going to be working with together in class-- 00:14:48.250 --> 00:14:53.730 align:middle line:84% is part of this folk music data set. 00:14:53.730 --> 00:14:57.520 align:middle line:84% And we're going to be looking at the Irish folk music corpus. 00:14:57.520 --> 00:15:01.270 align:middle line:84% There's a lot of Irish folk music corpora online, 00:15:01.270 --> 00:15:02.670 align:middle line:90% and we'll be looking at that. 00:15:02.670 --> 00:15:05.430 align:middle line:84% Then just to sum up, then we'll be 00:15:05.430 --> 00:15:08.010 align:middle line:84% looking at probably the most overlooked 00:15:08.010 --> 00:15:12.600 align:middle line:84% at corpus of all time, but one that there's 00:15:12.600 --> 00:15:16.020 align:middle line:84% a lot of written about it, and that is Johann Sebastian Bach's 00:15:16.020 --> 00:15:17.012 align:middle line:90% Chorales. 00:15:17.012 --> 00:15:18.970 align:middle line:84% The ones that you've already been working with, 00:15:18.970 --> 00:15:21.012 align:middle line:84% a couple of them, we're going to see what happens 00:15:21.012 --> 00:15:22.980 align:middle line:90% when we work with all of them. 00:15:22.980 --> 00:15:28.200 align:middle line:84% One of the things I'm going to be asking you to look at is one 00:15:28.200 --> 00:15:34.510 align:middle line:84% by next class, if you can, is to start reading an article by Bret 00:15:34.510 --> 00:15:37.220 align:middle line:84% Aarden and David Huron, which is-- 00:15:37.220 --> 00:15:41.990 align:middle line:84% as you can see from the quality of the scans-- 00:15:41.990 --> 00:15:43.400 align:middle line:90% a little bit of an old article. 00:15:43.400 --> 00:15:45.590 align:middle line:84% I think it's about 20 years old now. 00:15:45.590 --> 00:15:47.980 align:middle line:84% But it was one of the first ones to see 00:15:47.980 --> 00:15:51.520 align:middle line:84% what happens when you take musical features 00:15:51.520 --> 00:15:53.720 align:middle line:84% and compare them to geographic features. 00:15:53.720 --> 00:15:59.110 align:middle line:84% So looking at a density contour map of melodies 00:15:59.110 --> 00:16:01.850 align:middle line:84% where most phrases end on the tonic, 00:16:01.850 --> 00:16:05.950 align:middle line:84% we can see that there's a huge change from here to here in-- 00:16:05.950 --> 00:16:07.460 align:middle line:90% that's Denmark up here. 00:16:07.460 --> 00:16:10.850 align:middle line:84% This is Germany, right where the Alps began. 00:16:10.850 --> 00:16:14.290 align:middle line:84% And so we can see how geographic features 00:16:14.290 --> 00:16:19.220 align:middle line:90% might change how music is. 00:16:19.220 --> 00:16:22.310 align:middle line:84% So I'm going to blank that for a second. 00:16:22.310 --> 00:16:25.160 align:middle line:84% I'm going to open up Jupyter Notebook. 00:16:25.160 --> 00:16:28.910 align:middle line:84% And we're also going to unlock a few things along the way. 00:16:28.910 --> 00:16:31.700 align:middle line:84% So I'm hoping that we'll get through some 00:16:31.700 --> 00:16:36.260 align:middle line:84% of these Music21 topics as we go. 00:16:36.260 --> 00:16:42.590 align:middle line:84% So we'll start by importing something 00:16:42.590 --> 00:16:47.540 align:middle line:84% that we've already unlocked from Music21 import corpus. 00:16:47.540 --> 00:16:50.450 align:middle line:90% And we'll call IRL-- 00:16:50.450 --> 00:16:55.460 align:middle line:84% abbreviation for Ireland-- corpus parse. 00:16:55.460 --> 00:16:56.600 align:middle line:90% We'll be safe. 00:16:56.600 --> 00:17:02.360 align:middle line:84% Essen-- E-S-S-E-N-- folk song slash IRL. 00:17:02.360 --> 00:17:05.700 align:middle line:84% See how close I can get this to me so I'm still looking at you. 00:17:05.700 --> 00:17:06.200 align:middle line:90% Oops. 00:17:06.200 --> 00:17:09.920 align:middle line:90% 00:17:09.920 --> 00:17:14.839 align:middle line:90% Put this topic five, 2023. 00:17:14.839 --> 00:17:17.359 align:middle line:90% Great. 00:17:17.359 --> 00:17:19.849 align:middle line:84% So everybody, give me a quick thumbs up 00:17:19.849 --> 00:17:21.740 align:middle line:90% if we're at this place. 00:17:21.740 --> 00:17:22.295 align:middle line:90% Awesome. 00:17:22.295 --> 00:17:22.795 align:middle line:90% Super. 00:17:22.795 --> 00:17:26.438 align:middle line:90% 00:17:26.438 --> 00:17:26.938 align:middle line:90% OK. 00:17:26.938 --> 00:17:29.677 align:middle line:90% 00:17:29.677 --> 00:17:32.010 align:middle line:84% One of the things you can do, anything that you get from 00:17:32.010 --> 00:17:36.150 align:middle line:84% the corpus, just in case you ever need to look at it-- oops, 00:17:36.150 --> 00:17:40.440 align:middle line:90% sorry, I have an old thing-- 00:17:40.440 --> 00:17:48.060 align:middle line:84% is look at where it is on your hard drive. 00:17:48.060 --> 00:17:53.550 align:middle line:84% OK, now somebody can hack my hard drive, and so on. 00:17:53.550 --> 00:17:56.830 align:middle line:84% What kind of thing would you expect IRL to be? 00:17:56.830 --> 00:17:57.695 align:middle line:90% What kind of object? 00:17:57.695 --> 00:18:03.210 align:middle line:90% 00:18:03.210 --> 00:18:04.530 align:middle line:90% An int. 00:18:04.530 --> 00:18:07.090 align:middle line:90% Who votes for int? 00:18:07.090 --> 00:18:08.470 align:middle line:90% Float? 00:18:08.470 --> 00:18:08.970 align:middle line:90% String? 00:18:08.970 --> 00:18:16.710 align:middle line:90% 00:18:16.710 --> 00:18:17.670 align:middle line:90% You think it's string? 00:18:17.670 --> 00:18:18.120 align:middle line:90% No. 00:18:18.120 --> 00:18:18.828 align:middle line:90% You don't think-- 00:18:18.828 --> 00:18:21.840 align:middle line:84% AUDIENCE: No, I was going to say something else. 00:18:21.840 --> 00:18:26.290 align:middle line:84% MICHAEL CUTHBERT: What usually are we working with in Music21. 00:18:26.290 --> 00:18:27.148 align:middle line:90% AUDIENCE: Stream. 00:18:27.148 --> 00:18:28.190 align:middle line:90% MICHAEL CUTHBERT: Stream. 00:18:28.190 --> 00:18:28.690 align:middle line:90% Great. 00:18:28.690 --> 00:18:30.913 align:middle line:84% What kind of stream do we think this-- 00:18:30.913 --> 00:18:31.580 align:middle line:90% AUDIENCE: Score. 00:18:31.580 --> 00:18:32.580 align:middle line:90% MICHAEL CUTHBERT: Score. 00:18:32.580 --> 00:18:34.040 align:middle line:84% Score would be most logical, right? 00:18:34.040 --> 00:18:35.540 align:middle line:90% In this case, we're working-- 00:18:35.540 --> 00:18:38.950 align:middle line:84% this our first time working with something bigger than a score, 00:18:38.950 --> 00:18:40.220 align:middle line:90% something called an opus. 00:18:40.220 --> 00:18:42.560 align:middle line:90% So a collection of scores. 00:18:42.560 --> 00:18:45.220 align:middle line:84% And so what we can do is we can look 00:18:45.220 --> 00:18:51.290 align:middle line:90% at how many scores are in here. 00:18:51.290 --> 00:18:53.750 align:middle line:84% So this is a pretty small corpus, 00:18:53.750 --> 00:18:57.320 align:middle line:84% but big enough to start doing some statistics on. 00:18:57.320 --> 00:19:01.720 align:middle line:90% So we can do irl.scores0.show. 00:19:01.720 --> 00:19:08.950 align:middle line:90% 00:19:08.950 --> 00:19:14.060 align:middle line:90% And good deterministic. 00:19:14.060 --> 00:19:17.410 align:middle line:84% It's still the same one that I thought it would be. 00:19:17.410 --> 00:19:18.050 align:middle line:90% Let's see this. 00:19:18.050 --> 00:19:20.170 align:middle line:84% One of the things you'll find if you're 00:19:20.170 --> 00:19:24.620 align:middle line:84% really interested in being efficient 00:19:24.620 --> 00:19:29.540 align:middle line:84% is instead of doing what do you call it? 00:19:29.540 --> 00:19:32.120 align:middle line:90% Square bracket zero? 00:19:32.120 --> 00:19:37.440 align:middle line:84% On Music21 scores, once you've done all your filtering, 00:19:37.440 --> 00:19:39.680 align:middle line:84% you can do dot first and dot last. 00:19:39.680 --> 00:19:43.550 align:middle line:90% And all they are is O of one-- 00:19:43.550 --> 00:19:47.420 align:middle line:84% or that they stop finding all the scores 00:19:47.420 --> 00:19:48.720 align:middle line:90% once they found the first one. 00:19:48.720 --> 00:19:53.280 align:middle line:84% So it's a little bit-- if you're into algorithmic optimization, 00:19:53.280 --> 00:19:57.216 align:middle line:84% it can sometimes be quite a bit faster for very long notes. 00:19:57.216 --> 00:19:57.716 align:middle line:90% OK. 00:19:57.716 --> 00:20:01.688 align:middle line:90% So what I want to know is-- 00:20:01.688 --> 00:20:03.150 align:middle line:90% this is interesting. 00:20:03.150 --> 00:20:05.330 align:middle line:90% It's in 3/4. 00:20:05.330 --> 00:20:06.795 align:middle line:90% Pick up. 00:20:06.795 --> 00:20:07.295 align:middle line:90% Good. 00:20:07.295 --> 00:20:13.820 align:middle line:84% It does some of the things that we expect tonal pieces 00:20:13.820 --> 00:20:15.810 align:middle line:90% from folk era to do. 00:20:15.810 --> 00:20:18.230 align:middle line:90% If you begin with-- 00:20:18.230 --> 00:20:23.520 align:middle line:84% what is that-- a dotted eighth, a dotted quarter missing 00:20:23.520 --> 00:20:27.630 align:middle line:84% at the beginning or added to the pickup measure at the beginning, 00:20:27.630 --> 00:20:31.890 align:middle line:84% you expect that at the very end, you also have that. 00:20:31.890 --> 00:20:35.460 align:middle line:84% So the pickup measure and the ending measure 00:20:35.460 --> 00:20:38.182 align:middle line:90% usually represent one. 00:20:38.182 --> 00:20:38.890 align:middle line:90% Is everything OK? 00:20:38.890 --> 00:20:39.818 align:middle line:90% Yeah, go ahead. 00:20:39.818 --> 00:20:41.610 align:middle line:84% AUDIENCE: Could you go-- scroll up, please? 00:20:41.610 --> 00:20:43.027 align:middle line:84% MICHAEL CUTHBERT: Sure, of course. 00:20:43.027 --> 00:20:45.360 align:middle line:90% Always glad to. 00:20:45.360 --> 00:20:49.420 align:middle line:90% So did we get it? 00:20:49.420 --> 00:20:49.933 align:middle line:90% Is that-- 00:20:49.933 --> 00:20:50.850 align:middle line:90% AUDIENCE: That's good. 00:20:50.850 --> 00:20:52.470 align:middle line:84% Yeah, the piece looks a little bit different on mine. 00:20:52.470 --> 00:20:54.387 align:middle line:84% MICHAEL CUTHBERT: The piece looks a little bit 00:20:54.387 --> 00:20:56.850 align:middle line:90% different on yours? 00:20:56.850 --> 00:20:59.160 align:middle line:90% Hugely, terribly different? 00:20:59.160 --> 00:21:02.020 align:middle line:90% No, just a little bit different? 00:21:02.020 --> 00:21:04.350 align:middle line:90% OK. 00:21:04.350 --> 00:21:05.050 align:middle line:90% Yeah. 00:21:05.050 --> 00:21:09.420 align:middle line:84% I mean, depending on your version, 00:21:09.420 --> 00:21:12.750 align:middle line:84% here's what it looked like a couple of years ago. 00:21:12.750 --> 00:21:18.540 align:middle line:84% It's also quite a bit smaller back then and things, but OK. 00:21:18.540 --> 00:21:23.010 align:middle line:90% So it's in 3/4. 00:21:23.010 --> 00:21:24.965 align:middle line:84% Is that common for Irish folk music? 00:21:24.965 --> 00:21:28.810 align:middle line:90% 00:21:28.810 --> 00:21:30.370 align:middle line:90% What do you think? 00:21:30.370 --> 00:21:34.770 align:middle line:84% Who has heard Irish musicians play Irish music? 00:21:34.770 --> 00:21:36.690 align:middle line:90% Not very many. 00:21:36.690 --> 00:21:38.890 align:middle line:84% Keep your hand up if you're also over 21. 00:21:38.890 --> 00:21:40.500 align:middle line:90% No, just kidding. 00:21:40.500 --> 00:21:42.780 align:middle line:90% Because they do usually play-- 00:21:42.780 --> 00:21:45.060 align:middle line:84% one of the places you can hear quite often 00:21:45.060 --> 00:21:50.580 align:middle line:84% Irish folk musicians play in Boston is in pubs and things. 00:21:50.580 --> 00:21:53.190 align:middle line:84% So of people who have heard or people 00:21:53.190 --> 00:21:58.560 align:middle line:84% who think they might know this, what's a time signature that you 00:21:58.560 --> 00:22:00.130 align:middle line:90% think would be common? 00:22:00.130 --> 00:22:00.630 align:middle line:90% Adam? 00:22:00.630 --> 00:22:00.970 align:middle line:90% AUDIENCE: 2/4. 00:22:00.970 --> 00:22:01.740 align:middle line:90% MICHAEL CUTHBERT: 2/4. 00:22:01.740 --> 00:22:02.190 align:middle line:90% [? Sruthi? ?] 00:22:02.190 --> 00:22:02.610 align:middle line:90% AUDIENCE: 6/8? 00:22:02.610 --> 00:22:03.527 align:middle line:90% MICHAEL CUTHBERT: 6/8. 00:22:03.527 --> 00:22:04.180 align:middle line:90% John? 00:22:04.180 --> 00:22:04.785 align:middle line:90% 6/8 also? 00:22:04.785 --> 00:22:06.160 align:middle line:84% AUDIENCE: I was going to say 6/8. 00:22:06.160 --> 00:22:07.077 align:middle line:90% MICHAEL CUTHBERT: 6/8? 00:22:07.077 --> 00:22:10.140 align:middle line:84% Well, this is one of those cool questions where 00:22:10.140 --> 00:22:12.390 align:middle line:84% if you have the same answer, somebody else would still 00:22:12.390 --> 00:22:14.380 align:middle line:84% want to hear it because that starts building up consensus. 00:22:14.380 --> 00:22:15.790 align:middle line:84% Somebody give me a time signature 00:22:15.790 --> 00:22:18.950 align:middle line:84% you don't expect to be the highest-- 00:22:18.950 --> 00:22:22.100 align:middle line:84% the most commonly used time signature. 00:22:22.100 --> 00:22:22.739 align:middle line:90% Yeah, Jason? 00:22:22.739 --> 00:22:23.322 align:middle line:90% AUDIENCE: 7/4. 00:22:23.322 --> 00:22:24.680 align:middle line:90% MICHAEL CUTHBERT: 7/4. 00:22:24.680 --> 00:22:25.850 align:middle line:90% Good, good. 00:22:25.850 --> 00:22:28.240 align:middle line:90% OK, so we have some hypotheses. 00:22:28.240 --> 00:22:30.460 align:middle line:84% What we're going to do is try to figure out 00:22:30.460 --> 00:22:33.580 align:middle line:84% what's the most commonly used time signature 00:22:33.580 --> 00:22:35.480 align:middle line:90% in this Irish folk thing. 00:22:35.480 --> 00:22:37.250 align:middle line:84% And always start with a hypothesis 00:22:37.250 --> 00:22:41.200 align:middle line:84% because the number of times I've seen people go ahead and publish 00:22:41.200 --> 00:22:46.600 align:middle line:84% things where they don't have a hypothesis, 00:22:46.600 --> 00:22:48.310 align:middle line:90% and then they don't realize-- 00:22:48.310 --> 00:22:51.040 align:middle line:84% like, oh, I printed out the measure number 00:22:51.040 --> 00:22:55.660 align:middle line:84% divided by the number of notes, not the time signature. 00:22:55.660 --> 00:23:00.640 align:middle line:84% And therefore, 137 over 14,000 probably wasn't the most common 00:23:00.640 --> 00:23:01.370 align:middle line:90% time signature. 00:23:01.370 --> 00:23:03.010 align:middle line:90% So let's start. 00:23:03.010 --> 00:23:07.330 align:middle line:84% This was a question that came up at some point. 00:23:07.330 --> 00:23:11.860 align:middle line:84% Can you use-- are anything in the normal Python library 00:23:11.860 --> 00:23:12.880 align:middle line:90% unlocked in this class? 00:23:12.880 --> 00:23:13.380 align:middle line:90% Yes. 00:23:13.380 --> 00:23:15.170 align:middle line:90% So let's import the counter. 00:23:15.170 --> 00:23:17.640 align:middle line:90% Who has seen counter? 00:23:17.640 --> 00:23:18.540 align:middle line:90% Not everybody. 00:23:18.540 --> 00:23:19.970 align:middle line:90% OK, some people have. 00:23:19.970 --> 00:23:23.760 align:middle line:84% Counter is just-- it's a very cool little-- 00:23:23.760 --> 00:23:25.260 align:middle line:90% it's just a dictionary. 00:23:25.260 --> 00:23:28.890 align:middle line:84% But it starts off at zero when something's not there. 00:23:28.890 --> 00:23:30.230 align:middle line:90% So you don't have to say-- 00:23:30.230 --> 00:23:35.070 align:middle line:84% when you're adding things up, you can just start from that. 00:23:35.070 --> 00:23:39.920 align:middle line:84% So you can just start adding one to things. 00:23:39.920 --> 00:23:40.950 align:middle line:90% So I kind of like that. 00:23:40.950 --> 00:23:48.603 align:middle line:84% So let's say, for song in the Irish corpus, we can do-- 00:23:48.603 --> 00:23:50.520 align:middle line:84% we're going to want to get the time signature. 00:23:50.520 --> 00:23:53.836 align:middle line:84% So I want you to song dot something. 00:23:53.836 --> 00:23:56.750 align:middle line:90% 00:23:56.750 --> 00:23:57.320 align:middle line:90% Oops. 00:23:57.320 --> 00:23:59.990 align:middle line:90% Pull that up a little bit. 00:23:59.990 --> 00:24:01.950 align:middle line:90% And fill out the rest. 00:24:01.950 --> 00:24:04.410 align:middle line:84% Or maybe song-- maybe we don't even need the dot something. 00:24:04.410 --> 00:24:07.240 align:middle line:84% I don't know, song is probably the next thing. 00:24:07.240 --> 00:24:11.240 align:middle line:90% 00:24:11.240 --> 00:24:15.520 align:middle line:84% And we did put IRL scores first up there, 00:24:15.520 --> 00:24:17.135 align:middle line:84% so you know he can test this on one. 00:24:17.135 --> 00:24:19.830 align:middle line:90% 00:24:19.830 --> 00:24:22.773 align:middle line:84% Let's say we're only-- for simplicity's sake, 00:24:22.773 --> 00:24:24.690 align:middle line:84% we're only going to be counting the first time 00:24:24.690 --> 00:24:26.080 align:middle line:90% signature in each song. 00:24:26.080 --> 00:24:30.610 align:middle line:84% And that most of the songs in this collection-- 00:24:30.610 --> 00:24:32.530 align:middle line:84% I think all of them are monophonic, 00:24:32.530 --> 00:24:36.060 align:middle line:84% so you're only going to have one part. 00:24:36.060 --> 00:24:39.300 align:middle line:84% What are some words you're typing? 00:24:39.300 --> 00:24:40.600 align:middle line:90% AUDIENCE: Get element by class. 00:24:40.600 --> 00:24:41.558 align:middle line:90% MICHAEL CUTHBERT: Good. 00:24:41.558 --> 00:24:43.440 align:middle line:90% Get element by class, or get-- 00:24:43.440 --> 00:24:45.070 align:middle line:90% there's one little type. 00:24:45.070 --> 00:24:48.380 align:middle line:90% Get elements, plural. 00:24:48.380 --> 00:24:48.880 align:middle line:90% Yeah. 00:24:48.880 --> 00:24:53.500 align:middle line:84% So unfortunately, there isn't a first element thing. 00:24:53.500 --> 00:24:54.000 align:middle line:90% Great. 00:24:54.000 --> 00:24:57.030 align:middle line:84% Anybody else having other words that you're 00:24:57.030 --> 00:24:58.970 align:middle line:84% typing that you'd like everyone else to know? 00:24:58.970 --> 00:25:02.320 align:middle line:90% 00:25:02.320 --> 00:25:02.820 align:middle line:90% No? 00:25:02.820 --> 00:25:09.690 align:middle line:90% 00:25:09.690 --> 00:25:12.580 align:middle line:84% Unless I tell you otherwise, like the midterm or something, 00:25:12.580 --> 00:25:16.080 align:middle line:84% you're generally allowed to talk to people around you. 00:25:16.080 --> 00:25:26.080 align:middle line:90% 00:25:26.080 --> 00:25:28.750 align:middle line:84% So what's a word that you've typed? 00:25:28.750 --> 00:25:32.020 align:middle line:84% AUDIENCE: I'm going to just select the [INAUDIBLE]. 00:25:32.020 --> 00:25:35.200 align:middle line:90% MICHAEL CUTHBERT: OK, cool. 00:25:35.200 --> 00:25:36.080 align:middle line:90% Yeah, Misha? 00:25:36.080 --> 00:25:36.580 align:middle line:90% What's-- 00:25:36.580 --> 00:25:40.565 align:middle line:84% AUDIENCE: I guess we can, yeah, use meter.TimeSignature. 00:25:40.565 --> 00:25:42.440 align:middle line:84% MICHAEL CUTHBERT: OK, so meter.TimeSignature. 00:25:42.440 --> 00:25:45.050 align:middle line:84% Yeah, so you'll probably want to import meter. 00:25:45.050 --> 00:25:47.090 align:middle line:90% Sorry if I didn't do that. 00:25:47.090 --> 00:25:49.270 align:middle line:84% You can import anything that's unlocked always. 00:25:49.270 --> 00:25:52.213 align:middle line:90% So from Music21, import meter. 00:25:52.213 --> 00:25:53.630 align:middle line:84% And what's the word that you said? 00:25:53.630 --> 00:25:57.010 align:middle line:90% Meter.TimeSignature. 00:25:57.010 --> 00:25:58.575 align:middle line:90% That one might be helpful. 00:25:58.575 --> 00:26:03.850 align:middle line:90% 00:26:03.850 --> 00:26:07.380 align:middle line:84% Anybody else have any words that you've been typing? 00:26:07.380 --> 00:26:10.485 align:middle line:90% 00:26:10.485 --> 00:26:14.090 align:middle line:90% I'm going to type a word. 00:26:14.090 --> 00:26:15.890 align:middle line:90% My word is recurse. 00:26:15.890 --> 00:26:18.620 align:middle line:90% Anybody type that word? 00:26:18.620 --> 00:26:20.330 align:middle line:90% No? 00:26:20.330 --> 00:26:23.900 align:middle line:90% Or flatten. 00:26:23.900 --> 00:26:25.060 align:middle line:90% Anybody type that word? 00:26:25.060 --> 00:26:31.580 align:middle line:90% 00:26:31.580 --> 00:26:34.320 align:middle line:90% I'm only asking for this line. 00:26:34.320 --> 00:26:38.630 align:middle line:84% So I'm just trying to figure out who has gotten a time 00:26:38.630 --> 00:26:40.220 align:middle line:90% signature out? 00:26:40.220 --> 00:26:42.500 align:middle line:90% OK, so some people have. 00:26:42.500 --> 00:26:45.637 align:middle line:84% Vincent, what did you type to get your time? 00:26:45.637 --> 00:26:47.220 align:middle line:84% AUDIENCE: Oh, I made a get suggestion. 00:26:47.220 --> 00:26:49.175 align:middle line:90% So recurse.getElementsbyClass. 00:26:49.175 --> 00:26:50.300 align:middle line:90% MICHAEL CUTHBERT: OK, good. 00:26:50.300 --> 00:26:52.210 align:middle line:90% Recurse.getElementsbyClass. 00:26:52.210 --> 00:26:57.530 align:middle line:90% 00:26:57.530 --> 00:26:59.580 align:middle line:90% And anybody, help. 00:26:59.580 --> 00:27:02.960 align:middle line:84% What's the next thing I'm going to type? 00:27:02.960 --> 00:27:05.895 align:middle line:90% Meter.TimeSignature. 00:27:05.895 --> 00:27:06.395 align:middle line:90% Great. 00:27:06.395 --> 00:27:09.390 align:middle line:90% 00:27:09.390 --> 00:27:12.920 align:middle line:84% Anything else that I should be typing on that? 00:27:12.920 --> 00:27:16.290 align:middle line:90% 00:27:16.290 --> 00:27:18.130 align:middle line:90% Is this correct? 00:27:18.130 --> 00:27:23.040 align:middle line:84% Will TS now be a meter.TimeSignature? 00:27:23.040 --> 00:27:26.910 align:middle line:84% OK, let's try this on one song since I don't 00:27:26.910 --> 00:27:28.300 align:middle line:90% know if we have this at all. 00:27:28.300 --> 00:27:32.590 align:middle line:84% So we'll call-- we'll go back to the irl.scores.first. 00:27:32.590 --> 00:27:36.130 align:middle line:84% And what do we-- what was that name of that song? 00:27:36.130 --> 00:27:37.090 align:middle line:90% "A Lady Fair". 00:27:37.090 --> 00:27:44.160 align:middle line:84% We'll call it lady equals irl.scores.first. 00:27:44.160 --> 00:27:45.150 align:middle line:90% Now we'll try this. 00:27:45.150 --> 00:27:51.810 align:middle line:84% Lady.recurse.getElementsbyClass meter.TimeSignature. 00:27:51.810 --> 00:27:54.720 align:middle line:90% 00:27:54.720 --> 00:27:58.500 align:middle line:84% Who votes this is going to get me a time signature? 00:27:58.500 --> 00:28:00.990 align:middle line:90% Who votes not? 00:28:00.990 --> 00:28:01.858 align:middle line:90% OK. 00:28:01.858 --> 00:28:02.650 align:middle line:90% AUDIENCE: Not sure. 00:28:02.650 --> 00:28:03.900 align:middle line:90% MICHAEL CUTHBERT: Not sure. 00:28:03.900 --> 00:28:05.320 align:middle line:90% But not a time signature. 00:28:05.320 --> 00:28:07.350 align:middle line:84% You had your hand up on not a time signature? 00:28:07.350 --> 00:28:07.850 align:middle line:90% Good. 00:28:07.850 --> 00:28:08.778 align:middle line:90% Anybody else? 00:28:08.778 --> 00:28:10.070 align:middle line:90% Vanessa, you look like you're-- 00:28:10.070 --> 00:28:12.100 align:middle line:90% AUDIENCE: [INAUDIBLE]. 00:28:12.100 --> 00:28:13.350 align:middle line:90% MICHAEL CUTHBERT: It's a what? 00:28:13.350 --> 00:28:15.630 align:middle line:84% AUDIENCE: I think it's like a getter function [INAUDIBLE]. 00:28:15.630 --> 00:28:17.870 align:middle line:84% MICHAEL CUTHBERT: Yeah, something that will get what? 00:28:17.870 --> 00:28:20.180 align:middle line:90% AUDIENCE: [INAUDIBLE]. 00:28:20.180 --> 00:28:23.962 align:middle line:84% From the song as opposed to the song element itself. 00:28:23.962 --> 00:28:24.920 align:middle line:90% MICHAEL CUTHBERT: Oops. 00:28:24.920 --> 00:28:27.290 align:middle line:90% Do I not have-- 00:28:27.290 --> 00:28:28.920 align:middle line:90% oh, sorry. 00:28:28.920 --> 00:28:32.750 align:middle line:84% Yes, that won't return anything if I don't put these, 00:28:32.750 --> 00:28:34.395 align:middle line:90% brackets right? 00:28:34.395 --> 00:28:36.020 align:middle line:84% As I was trying to do it on a function. 00:28:36.020 --> 00:28:39.570 align:middle line:90% 00:28:39.570 --> 00:28:46.580 align:middle line:84% Oh, this comes up, by the way, in problem sets sometimes 00:28:46.580 --> 00:28:47.330 align:middle line:90% in answers. 00:28:47.330 --> 00:28:48.450 align:middle line:90% The order matters, right? 00:28:48.450 --> 00:28:52.910 align:middle line:84% I hadn't run the Music21 import meter yet. 00:28:52.910 --> 00:28:55.140 align:middle line:90% Just something to keep in mind. 00:28:55.140 --> 00:28:56.880 align:middle line:84% Always make sure when you're done-- 00:28:56.880 --> 00:28:58.547 align:middle line:84% when you think your problem set is done, 00:28:58.547 --> 00:29:01.530 align:middle line:84% hit Restart and Run All, and see if it gets to the end 00:29:01.530 --> 00:29:04.220 align:middle line:84% because you might have been solving a problem along the way 00:29:04.220 --> 00:29:05.530 align:middle line:90% and then have-- 00:29:05.530 --> 00:29:09.170 align:middle line:84% oh, in order to solve question two, 00:29:09.170 --> 00:29:10.940 align:middle line:84% I imported something from question three, 00:29:10.940 --> 00:29:13.010 align:middle line:84% but then it's not ready when I run it. 00:29:13.010 --> 00:29:15.560 align:middle line:84% So yeah, so now that we have this, 00:29:15.560 --> 00:29:21.860 align:middle line:84% what is something called a recursive iterator? 00:29:21.860 --> 00:29:24.190 align:middle line:90% So it's not a time signature. 00:29:24.190 --> 00:29:29.440 align:middle line:84% How do I get the time signature from this if we have this? 00:29:29.440 --> 00:29:30.920 align:middle line:90% Anybody? 00:29:30.920 --> 00:29:31.420 align:middle line:90% Hannah? 00:29:31.420 --> 00:29:33.040 align:middle line:90% AUDIENCE: You can index zero. 00:29:33.040 --> 00:29:35.440 align:middle line:90% MICHAEL CUTHBERT: Index zero. 00:29:35.440 --> 00:29:36.460 align:middle line:90% Perfect. 00:29:36.460 --> 00:29:41.110 align:middle line:84% It's kind of like what we did here, indexing 00:29:41.110 --> 00:29:44.352 align:middle line:90% zero iterator over scores. 00:29:44.352 --> 00:29:46.060 align:middle line:84% But now we have something that's indexing 00:29:46.060 --> 00:29:47.240 align:middle line:90% over the time signature. 00:29:47.240 --> 00:29:50.080 align:middle line:84% So we should be able to do this with-- 00:29:50.080 --> 00:29:53.620 align:middle line:84% if I remember my parentheses, I do 00:29:53.620 --> 00:29:57.670 align:middle line:84% get used to an IDE that fills in my parentheses for me, which 00:29:57.670 --> 00:29:58.430 align:middle line:90% I'm not using. 00:29:58.430 --> 00:30:04.760 align:middle line:84% And I can do index zero or I can do dot first, either one. 00:30:04.760 --> 00:30:07.370 align:middle line:84% This is a particular case where dot first 00:30:07.370 --> 00:30:09.303 align:middle line:84% will save you a lot of time because you're 00:30:09.303 --> 00:30:10.970 align:middle line:84% pretty sure the time signature should be 00:30:10.970 --> 00:30:12.810 align:middle line:90% at the beginning of the score. 00:30:12.810 --> 00:30:14.623 align:middle line:90% So once you hit one, stop. 00:30:14.623 --> 00:30:16.790 align:middle line:84% Don't go through the whole rest of the score looking 00:30:16.790 --> 00:30:19.600 align:middle line:84% for more time signatures if you only want the first one. 00:30:19.600 --> 00:30:20.100 align:middle line:90% Good. 00:30:20.100 --> 00:30:22.010 align:middle line:90% So we have the time signature. 00:30:22.010 --> 00:30:24.950 align:middle line:84% Now what we can do-- this is something we haven't seen. 00:30:24.950 --> 00:30:29.180 align:middle line:84% The ratio of a time signature-- the 3/4 sign-- 00:30:29.180 --> 00:30:31.260 align:middle line:90% is in its ratio string. 00:30:31.260 --> 00:30:33.750 align:middle line:84% There's also a numerator and denominator, 00:30:33.750 --> 00:30:36.080 align:middle line:84% but this is the one that's pretty helpful. 00:30:36.080 --> 00:30:42.900 align:middle line:84% And then meter counter ratio, what should I be doing now 00:30:42.900 --> 00:30:43.900 align:middle line:90% if I want to count them? 00:30:43.900 --> 00:30:47.747 align:middle line:90% 00:30:47.747 --> 00:30:48.830 align:middle line:90% AUDIENCE: Plus equals one? 00:30:48.830 --> 00:30:50.250 align:middle line:84% MICHAEL CUTHBERT: Plus equals one. 00:30:50.250 --> 00:30:51.590 align:middle line:90% Good. 00:30:51.590 --> 00:30:54.835 align:middle line:84% And then at the very end, let's see our meter counter. 00:30:54.835 --> 00:30:57.530 align:middle line:90% 00:30:57.530 --> 00:31:02.850 align:middle line:84% So we had hypotheses of 3/4 and 6/8. 00:31:02.850 --> 00:31:06.630 align:middle line:90% I put my money on 6/8. 00:31:06.630 --> 00:31:08.050 align:middle line:90% And whoops, what did I forget? 00:31:08.050 --> 00:31:11.970 align:middle line:90% 00:31:11.970 --> 00:31:14.790 align:middle line:90% Oh, irl.scores. 00:31:14.790 --> 00:31:16.830 align:middle line:90% Thank you. 00:31:16.830 --> 00:31:21.360 align:middle line:84% Irl.scores because the first thing there was metadata. 00:31:21.360 --> 00:31:33.060 align:middle line:90% 00:31:33.060 --> 00:31:34.742 align:middle line:90% So what's our winner? 00:31:34.742 --> 00:31:35.325 align:middle line:90% AUDIENCE: 3/4. 00:31:35.325 --> 00:31:36.510 align:middle line:90% MICHAEL CUTHBERT: 3/4. 00:31:36.510 --> 00:31:37.980 align:middle line:90% Our runner up? 00:31:37.980 --> 00:31:38.688 align:middle line:90% AUDIENCE: 4/4. 00:31:38.688 --> 00:31:42.580 align:middle line:84% MICHAEL CUTHBERT: 4/4 And then 6/8. 00:31:42.580 --> 00:31:45.670 align:middle line:84% At least 6/8 was on-- those of us who were 6/8, the 6/8 crew, 00:31:45.670 --> 00:31:47.290 align:middle line:90% we did OK on there. 00:31:47.290 --> 00:31:49.020 align:middle line:84% But yeah, I might have the difference 00:31:49.020 --> 00:31:52.180 align:middle line:84% between this being sung music versus fiddle music, 00:31:52.180 --> 00:31:52.810 align:middle line:90% I don't know. 00:31:52.810 --> 00:31:56.590 align:middle line:84% We can try some of the other things in there. 00:31:56.590 --> 00:31:58.650 align:middle line:84% By the way, with a counter, you can also 00:31:58.650 --> 00:32:01.960 align:middle line:84% do things like meter counter.most 00:32:01.960 --> 00:32:08.470 align:middle line:84% common, three most common, and see what they are in that order. 00:32:08.470 --> 00:32:10.060 align:middle line:90% Cool. 00:32:10.060 --> 00:32:16.810 align:middle line:84% So that's one of the things we can do with this corpus. 00:32:16.810 --> 00:32:22.430 align:middle line:84% Let's jump over to the Bach corpus for now for a little bit. 00:32:22.430 --> 00:32:24.320 align:middle line:90% Are there questions about this? 00:32:24.320 --> 00:32:26.750 align:middle line:84% I mean, that's just the tiniest little thing. 00:32:26.750 --> 00:32:29.050 align:middle line:84% We can figure out all the questions 00:32:29.050 --> 00:32:30.830 align:middle line:90% that we had before with it. 00:32:30.830 --> 00:32:33.760 align:middle line:90% Oh, here. 00:32:33.760 --> 00:32:35.440 align:middle line:90% Let's do this. 00:32:35.440 --> 00:32:40.720 align:middle line:84% Let's do this type of experiment that Sophia Son 00:32:40.720 --> 00:32:46.930 align:middle line:84% was talking about with position on the beat or whatever. 00:32:46.930 --> 00:32:54.430 align:middle line:84% Let's look at where notes with accidentals appear. 00:32:54.430 --> 00:33:00.650 align:middle line:84% Are they most commonly in this little passing position 00:33:00.650 --> 00:33:03.260 align:middle line:90% in the middle of the bar? 00:33:03.260 --> 00:33:06.710 align:middle line:84% Or are they more often contradicting right 00:33:06.710 --> 00:33:09.710 align:middle line:90% at the downbeat? 00:33:09.710 --> 00:33:11.390 align:middle line:90% Hypothesis? 00:33:11.390 --> 00:33:14.770 align:middle line:84% Who votes for accidentals more in the middle of the bar? 00:33:14.770 --> 00:33:18.290 align:middle line:90% 00:33:18.290 --> 00:33:18.950 align:middle line:90% Good. 00:33:18.950 --> 00:33:22.850 align:middle line:90% 7, 8-- and for on the downbeat? 00:33:22.850 --> 00:33:24.020 align:middle line:90% Good. 00:33:24.020 --> 00:33:26.840 align:middle line:90% Who's not participating? 00:33:26.840 --> 00:33:28.790 align:middle line:84% That didn't add up to-- let's one more time. 00:33:28.790 --> 00:33:32.300 align:middle line:90% In the middle of the bar? 00:33:32.300 --> 00:33:33.835 align:middle line:90% Or downbeat? 00:33:33.835 --> 00:33:34.710 align:middle line:90% A couple of downbeat. 00:33:34.710 --> 00:33:35.210 align:middle line:90% Good. 00:33:35.210 --> 00:33:36.900 align:middle line:84% I don't know what the answer to this is. 00:33:36.900 --> 00:33:39.960 align:middle line:84% Always just be thinking about these things. 00:33:39.960 --> 00:33:43.310 align:middle line:84% So let's do the same type of thing. 00:33:43.310 --> 00:33:45.510 align:middle line:90% We'll create another counter. 00:33:45.510 --> 00:33:49.460 align:middle line:84% And we'll unlock another little thing for you. 00:33:49.460 --> 00:33:56.890 align:middle line:90% So we'll call this accidental-- 00:33:56.890 --> 00:33:59.790 align:middle line:84% we're not counting accidentals, so maybe 00:33:59.790 --> 00:34:03.595 align:middle line:90% just accidental positions. 00:34:03.595 --> 00:34:05.470 align:middle line:84% But maybe we don't even want to do a counter. 00:34:05.470 --> 00:34:12.840 align:middle line:84% Maybe let's just do a list so we can run averages on it. 00:34:12.840 --> 00:34:16.320 align:middle line:90% Now, we'll be fair to compare-- 00:34:16.320 --> 00:34:17.820 align:middle line:90% well, I don't know-- 00:34:17.820 --> 00:34:19.810 align:middle line:90% where in the measure it is. 00:34:19.810 --> 00:34:23.670 align:middle line:84% Maybe we want to just say we'll just look at the 21 3/4 one 00:34:23.670 --> 00:34:26.620 align:middle line:84% so that all of our measures are the same length, 00:34:26.620 --> 00:34:30.750 align:middle line:84% except for, hopefully, the pickup measure and the end 00:34:30.750 --> 00:34:34.260 align:middle line:84% measure are rounding errors, but maybe not. 00:34:34.260 --> 00:34:37.150 align:middle line:84% You know, because they don't have the same length. 00:34:37.150 --> 00:34:42.810 align:middle line:84% So what we can do is, again, copy and paste. 00:34:42.810 --> 00:34:45.489 align:middle line:84% Probably, if we're doing this more, we can get this. 00:34:45.489 --> 00:34:52.050 align:middle line:84% And instead, we want to look at every note. 00:34:52.050 --> 00:34:56.650 align:middle line:84% So there's something called notes on this. 00:34:56.650 --> 00:34:59.440 align:middle line:84% Or actually, we'll do even better. 00:34:59.440 --> 00:35:01.270 align:middle line:84% We haven't imported note yet, right? 00:35:01.270 --> 00:35:08.170 align:middle line:90% From Music21, import note. 00:35:08.170 --> 00:35:13.540 align:middle line:84% Because of a bad decision I made a while back, 00:35:13.540 --> 00:35:15.927 align:middle line:84% dot notes also will get you chords 00:35:15.927 --> 00:35:17.260 align:middle line:90% if there are any chords in here. 00:35:17.260 --> 00:35:19.137 align:middle line:90% And chords, we will get to soon. 00:35:19.137 --> 00:35:20.720 align:middle line:84% They have to be dealt with separately. 00:35:20.720 --> 00:35:23.155 align:middle line:84% So we want to say .recurse.getElem 00:35:23.155 --> 00:35:25.070 align:middle line:90% entsbyClassnote.Note. 00:35:25.070 --> 00:35:27.170 align:middle line:84% But there's another way to do that. 00:35:27.170 --> 00:35:30.140 align:middle line:84% Anyone remember what's the shortcut for recurse? 00:35:30.140 --> 00:35:32.950 align:middle line:90% Get elements by class? 00:35:32.950 --> 00:35:34.120 align:middle line:90% Yeah, go ahead. 00:35:34.120 --> 00:35:36.492 align:middle line:84% AUDIENCE: You use the bracket and the class name. 00:35:36.492 --> 00:35:37.450 align:middle line:90% MICHAEL CUTHBERT: Good. 00:35:37.450 --> 00:35:44.330 align:middle line:84% So we'll put the bracket for n in song note.Note. 00:35:44.330 --> 00:35:47.150 align:middle line:84% So that will get us every note there. 00:35:47.150 --> 00:35:53.810 align:middle line:84% And how do we filter out ones without accidentals? 00:35:53.810 --> 00:35:55.560 align:middle line:90% So try that yourself. 00:35:55.560 --> 00:35:57.200 align:middle line:90% Try to figure out-- 00:35:57.200 --> 00:36:00.870 align:middle line:84% pretty sure that this is right for now. 00:36:00.870 --> 00:36:01.460 align:middle line:90% Yep. 00:36:01.460 --> 00:36:02.880 align:middle line:84% We've got a whole bunch of notes. 00:36:02.880 --> 00:36:07.850 align:middle line:84% We want to filter out ones without accidentals. 00:36:07.850 --> 00:36:10.340 align:middle line:84% Oh, why do we have a five space thing? 00:36:10.340 --> 00:36:13.550 align:middle line:90% That bothers me. 00:36:13.550 --> 00:36:16.926 align:middle line:90% Filter out no accidentals. 00:36:16.926 --> 00:36:24.140 align:middle line:90% 00:36:24.140 --> 00:36:25.345 align:middle line:90% We're going to have some-- 00:36:25.345 --> 00:36:28.940 align:middle line:90% 00:36:28.940 --> 00:36:30.923 align:middle line:84% we might not get to the Bach example 00:36:30.923 --> 00:36:33.590 align:middle line:84% because I came up with this one, and there's a neat little twist 00:36:33.590 --> 00:36:36.920 align:middle line:90% that I had forgotten about. 00:36:36.920 --> 00:36:40.490 align:middle line:84% Go ahead and interrupt the class, 00:36:40.490 --> 00:36:43.310 align:middle line:84% and ask that for everybody in a loud voice. 00:36:43.310 --> 00:36:47.150 align:middle line:84% AUDIENCE: What do people think about accidentals 00:36:47.150 --> 00:36:48.180 align:middle line:90% in the key signature? 00:36:48.180 --> 00:36:50.970 align:middle line:84% Are we counting those as accidentals? 00:36:50.970 --> 00:36:52.915 align:middle line:90% AUDIENCE: No. 00:36:52.915 --> 00:36:54.040 align:middle line:90% MICHAEL CUTHBERT: OK, good. 00:36:54.040 --> 00:36:56.040 align:middle line:84% So we're going to have to do some-- that was 00:36:56.040 --> 00:36:57.370 align:middle line:90% the wrinkle that I thought of. 00:36:57.370 --> 00:36:59.750 align:middle line:84% One of the ways you could do is you could say, oh shoot. 00:36:59.750 --> 00:37:00.250 align:middle line:90% OK. 00:37:00.250 --> 00:37:05.380 align:middle line:84% We're just going to look at ones with C major so we can do it. 00:37:05.380 --> 00:37:08.370 align:middle line:84% But presumably, sometimes you can filter down your corpus 00:37:08.370 --> 00:37:10.650 align:middle line:84% to get rid of some corner cases that 00:37:10.650 --> 00:37:13.810 align:middle line:84% might be a little bit too hard to deal with, 00:37:13.810 --> 00:37:15.550 align:middle line:90% at least in the first run. 00:37:15.550 --> 00:37:17.640 align:middle line:84% But getting rid of everything but C major 00:37:17.640 --> 00:37:20.530 align:middle line:84% is probably not going to leave us with enough of a corpus. 00:37:20.530 --> 00:37:26.400 align:middle line:84% So let's temporarily unlock a second module 00:37:26.400 --> 00:37:28.643 align:middle line:90% from Music21 import key. 00:37:28.643 --> 00:37:31.170 align:middle line:90% 00:37:31.170 --> 00:37:37.290 align:middle line:84% And first, let's get the first key signature for each thing 00:37:37.290 --> 00:37:38.225 align:middle line:90% so we can do-- 00:37:38.225 --> 00:37:41.400 align:middle line:90% 00:37:41.400 --> 00:37:43.920 align:middle line:84% how do we get the first key signature in the piece? 00:37:43.920 --> 00:37:45.750 align:middle line:84% Let's assume that for now, that it 00:37:45.750 --> 00:37:47.316 align:middle line:90% doesn't switch key signatures. 00:37:47.316 --> 00:37:52.110 align:middle line:90% 00:37:52.110 --> 00:37:54.870 align:middle line:84% AUDIENCE: Song.recurse parentheses. 00:37:54.870 --> 00:37:57.930 align:middle line:84% MICHAEL CUTHBERT: Song.recurse parentheses.getElementsbyClass. 00:37:57.930 --> 00:38:00.810 align:middle line:90% 00:38:00.810 --> 00:38:03.398 align:middle line:90% Or what's a simpler way? 00:38:03.398 --> 00:38:04.440 align:middle line:90% AUDIENCE: Square bracket. 00:38:04.440 --> 00:38:07.050 align:middle line:84% MICHAEL CUTHBERT: Square bracket key dot-- 00:38:07.050 --> 00:38:10.064 align:middle line:84% I'll tell you it's key signature-- 00:38:10.064 --> 00:38:10.920 align:middle line:90% dot first. 00:38:10.920 --> 00:38:14.820 align:middle line:90% 00:38:14.820 --> 00:38:18.600 align:middle line:84% So first, let's filter out everything-- 00:38:18.600 --> 00:38:22.660 align:middle line:84% so now, let me make a little introduction. 00:38:22.660 --> 00:38:24.700 align:middle line:84% Man, I wish I could edit at this part. 00:38:24.700 --> 00:38:26.500 align:middle line:90% So let's go back. 00:38:26.500 --> 00:38:28.540 align:middle line:90% Always work through one score. 00:38:28.540 --> 00:38:35.010 align:middle line:84% So we have the Lady score, and we'll get ks equals lady-- 00:38:35.010 --> 00:38:36.870 align:middle line:90% was it "My Lady Fair"? 00:38:36.870 --> 00:38:43.385 align:middle line:84% Lady key-- I didn't import key yet from Music21. 00:38:43.385 --> 00:38:46.080 align:middle line:90% 00:38:46.080 --> 00:38:49.390 align:middle line:84% Let's pause for a little bit and help 00:38:49.390 --> 00:38:52.640 align:middle line:84% me get caught up so that anybody watching this later 00:38:52.640 --> 00:38:53.390 align:middle line:90% can get caught up. 00:38:53.390 --> 00:39:00.640 align:middle line:84% Can you all come with me for a second and eyes for a second? 00:39:00.640 --> 00:39:01.900 align:middle line:90% Thanks. 00:39:01.900 --> 00:39:06.520 align:middle line:84% Can somebody tell me what I should be putting here? 00:39:06.520 --> 00:39:09.590 align:middle line:84% I think we're mostly at about this line together. 00:39:09.590 --> 00:39:15.790 align:middle line:84% So who have I not heard from very much? 00:39:15.790 --> 00:39:17.925 align:middle line:90% Angelica? 00:39:17.925 --> 00:39:19.730 align:middle line:84% AUDIENCE: OK, so one doesn't work. 00:39:19.730 --> 00:39:24.430 align:middle line:84% But I tried to say if n.pitch.accidental, 00:39:24.430 --> 00:39:29.740 align:middle line:84% and if the key signature.accidental by step, 00:39:29.740 --> 00:39:31.660 align:middle line:90% and then n.name. 00:39:31.660 --> 00:39:38.180 align:middle line:84% MICHAEL CUTHBERT: And-- sorry, and ks.accidental by step, 00:39:38.180 --> 00:39:38.680 align:middle line:90% n.name? 00:39:38.680 --> 00:39:43.750 align:middle line:90% 00:39:43.750 --> 00:39:44.320 align:middle line:90% Good. 00:39:44.320 --> 00:39:45.500 align:middle line:90% That's a pretty good start. 00:39:45.500 --> 00:39:47.220 align:middle line:90% I might change one thing on it. 00:39:47.220 --> 00:39:47.940 align:middle line:90% Anybody else? 00:39:47.940 --> 00:39:52.850 align:middle line:84% Anyone have a suggestion of one thing to change on that? 00:39:52.850 --> 00:39:55.520 align:middle line:84% So we're looking to see there is an accidental 00:39:55.520 --> 00:39:57.680 align:middle line:84% and there is an accidental-- yeah. 00:39:57.680 --> 00:39:59.360 align:middle line:84% AUDIENCE: I would change name to step. 00:39:59.360 --> 00:40:00.943 align:middle line:84% MICHAEL CUTHBERT: Change name to step. 00:40:00.943 --> 00:40:04.070 align:middle line:84% So the difference between name and step, accidental by step 00:40:04.070 --> 00:40:05.270 align:middle line:90% expects a step. 00:40:05.270 --> 00:40:09.890 align:middle line:90% C sharp's name is C sharp. 00:40:09.890 --> 00:40:11.510 align:middle line:90% C sharp step is-- 00:40:11.510 --> 00:40:14.270 align:middle line:90% shout it out. 00:40:14.270 --> 00:40:15.450 align:middle line:90% Nobody knows. 00:40:15.450 --> 00:40:16.190 align:middle line:90% Total mystery. 00:40:16.190 --> 00:40:17.570 align:middle line:90% C. Good. 00:40:17.570 --> 00:40:23.340 align:middle line:90% Step-- I know you knew that. 00:40:23.340 --> 00:40:24.720 align:middle line:90% But just keep that going. 00:40:24.720 --> 00:40:25.220 align:middle line:90% Good. 00:40:25.220 --> 00:40:28.440 align:middle line:90% So what's this logic? 00:40:28.440 --> 00:40:34.340 align:middle line:84% If the note pitch has an accidental and there 00:40:34.340 --> 00:40:38.240 align:middle line:84% is an accidental there, then what? 00:40:38.240 --> 00:40:40.070 align:middle line:84% Anybody who feels confident that they 00:40:40.070 --> 00:40:46.240 align:middle line:84% can code the next line of this, even if you didn't have this? 00:40:46.240 --> 00:40:49.440 align:middle line:84% Or even just logically, what do we want to do now? 00:40:49.440 --> 00:40:52.198 align:middle line:90% 00:40:52.198 --> 00:40:53.490 align:middle line:90% AUDIENCE: If they're the same-- 00:40:53.490 --> 00:40:54.280 align:middle line:90% MICHAEL CUTHBERT: Uh huh. 00:40:54.280 --> 00:40:55.630 align:middle line:90% AUDIENCE: --we don't add one. 00:40:55.630 --> 00:40:55.975 align:middle line:90% MICHAEL CUTHBERT: OK. 00:40:55.975 --> 00:40:57.590 align:middle line:84% If they're the same, we don't add one. 00:40:57.590 --> 00:40:58.300 align:middle line:90% Good. 00:40:58.300 --> 00:41:09.190 align:middle line:84% If n.pitch.accidental and equals ks.accidental by step, n.step-- 00:41:09.190 --> 00:41:12.940 align:middle line:90% 00:41:12.940 --> 00:41:13.670 align:middle line:90% what is it? 00:41:13.670 --> 00:41:16.130 align:middle line:90% Something plus equals zero. 00:41:16.130 --> 00:41:17.120 align:middle line:90% We won't do anything. 00:41:17.120 --> 00:41:18.850 align:middle line:90% Nil, we can pass. 00:41:18.850 --> 00:41:21.250 align:middle line:84% Since we want to do something at this point, 00:41:21.250 --> 00:41:24.280 align:middle line:90% how can we flip those? 00:41:24.280 --> 00:41:27.460 align:middle line:84% How can we flip it so we can add this? 00:41:27.460 --> 00:41:31.035 align:middle line:84% Don't worry, we'll improve some of the redundancy in a second. 00:41:31.035 --> 00:41:36.220 align:middle line:90% 00:41:36.220 --> 00:41:38.360 align:middle line:84% If it's like this, we don't want to do anything. 00:41:38.360 --> 00:41:40.690 align:middle line:84% If it's not like this, what do we want to do? 00:41:40.690 --> 00:41:41.780 align:middle line:90% Add one. 00:41:41.780 --> 00:41:44.690 align:middle line:90% So how can we save ourselves? 00:41:44.690 --> 00:41:50.170 align:middle line:84% I suppose we can put else something plus equals one. 00:41:50.170 --> 00:41:50.980 align:middle line:90% But-- 00:41:50.980 --> 00:41:52.570 align:middle line:90% AUDIENCE: You could do a not. 00:41:52.570 --> 00:41:53.820 align:middle line:90% MICHAEL CUTHBERT: I can do a-- 00:41:53.820 --> 00:41:54.700 align:middle line:90% AUDIENCE: A not. 00:41:54.700 --> 00:41:55.360 align:middle line:90% MICHAEL CUTHBERT: A not. 00:41:55.360 --> 00:41:55.940 align:middle line:90% Not equal. 00:41:55.940 --> 00:42:00.080 align:middle line:84% If these are not equal, then we can do that. 00:42:00.080 --> 00:42:01.970 align:middle line:90% So that could work. 00:42:01.970 --> 00:42:10.750 align:middle line:84% What's another thing I can probably save some redundancy? 00:42:10.750 --> 00:42:14.725 align:middle line:84% Let's store these things at acc equals n.pitch.accidental. 00:42:14.725 --> 00:42:18.064 align:middle line:90% 00:42:18.064 --> 00:42:23.020 align:middle line:90% ks acc equals this. 00:42:23.020 --> 00:42:27.770 align:middle line:84% When you find yourself writing something twice in two lines, 00:42:27.770 --> 00:42:29.685 align:middle line:84% you will probably continue to write it. 00:42:29.685 --> 00:42:37.000 align:middle line:90% 00:42:37.000 --> 00:42:39.750 align:middle line:90% And so we can do this. 00:42:39.750 --> 00:42:42.140 align:middle line:84% We might even be able to reduce this 00:42:42.140 --> 00:42:44.060 align:middle line:84% further because Python will allow 00:42:44.060 --> 00:42:53.030 align:middle line:84% you to compare none not equal to something else, but good. 00:42:53.030 --> 00:42:54.720 align:middle line:90% Does that cover all the cases? 00:42:54.720 --> 00:42:59.475 align:middle line:84% We have an accidental and we have the accidental 00:42:59.475 --> 00:43:00.350 align:middle line:90% in the key signature. 00:43:00.350 --> 00:43:01.892 align:middle line:84% Angelica, by the way, that was great. 00:43:01.892 --> 00:43:05.540 align:middle line:84% Thank you for getting us going on the right path. 00:43:05.540 --> 00:43:07.950 align:middle line:84% And we have a key signature accidental, 00:43:07.950 --> 00:43:10.260 align:middle line:84% and so we have an accident on the note. 00:43:10.260 --> 00:43:11.850 align:middle line:84% We have a key signature accidental. 00:43:11.850 --> 00:43:14.280 align:middle line:84% And the two are not equal to each other, 00:43:14.280 --> 00:43:17.795 align:middle line:84% so therefore, we're going to add one to something. 00:43:17.795 --> 00:43:19.920 align:middle line:84% We'll figure out what we're going to add something, 00:43:19.920 --> 00:43:21.510 align:middle line:84% but we care about that note, right? 00:43:21.510 --> 00:43:24.560 align:middle line:90% Any other notes we care about? 00:43:24.560 --> 00:43:25.160 align:middle line:90% Yeah? 00:43:25.160 --> 00:43:27.410 align:middle line:84% AUDIENCE: What about notes where the key signature has 00:43:27.410 --> 00:43:30.030 align:middle line:84% them sharped or flatted, But, then we natural them? 00:43:30.030 --> 00:43:31.300 align:middle line:90% MICHAEL CUTHBERT: OK, good. 00:43:31.300 --> 00:43:31.800 align:middle line:90% Good. 00:43:31.800 --> 00:43:36.843 align:middle line:84% And the problem with natural is what? 00:43:36.843 --> 00:43:38.760 align:middle line:84% AUDIENCE: Technically, it's not an accidental. 00:43:38.760 --> 00:43:41.400 align:middle line:84% MICHAEL CUTHBERT: Technically, it's not an accidental. 00:43:41.400 --> 00:43:45.390 align:middle line:84% This is a thing that I did to speed up so that there would 00:43:45.390 --> 00:43:50.680 align:middle line:84% be one fewer object creation back in 2004 and 2005, 00:43:50.680 --> 00:43:52.830 align:middle line:84% when object creation was still something 00:43:52.830 --> 00:43:54.340 align:middle line:90% that was slow in Python. 00:43:54.340 --> 00:43:57.840 align:middle line:84% Do I regret not putting an accidental of natural 00:43:57.840 --> 00:43:59.770 align:middle line:90% on every single note now? 00:43:59.770 --> 00:44:02.920 align:middle line:90% Yes, I do regret that. 00:44:02.920 --> 00:44:07.580 align:middle line:84% Here's a trick that I've started to do a lot since I don't really 00:44:07.580 --> 00:44:08.080 align:middle line:90% care. 00:44:08.080 --> 00:44:12.330 align:middle line:90% We'll import pitch for a second. 00:44:12.330 --> 00:44:20.970 align:middle line:84% Music21 import pitch, and we'll just make one natural. 00:44:20.970 --> 00:44:23.850 align:middle line:84% And I'll put it in all caps because that means to me 00:44:23.850 --> 00:44:29.370 align:middle line:84% that I'm going to try as much as possible never to mutate this. 00:44:29.370 --> 00:44:36.070 align:middle line:84% So one of the things we could do now is since none returns 00:44:36.070 --> 00:44:37.130 align:middle line:90% falsy-- 00:44:37.130 --> 00:44:39.460 align:middle line:84% so therefore, we continue in the or-- we 00:44:39.460 --> 00:44:42.550 align:middle line:84% can say the accidental is this one or natural, 00:44:42.550 --> 00:44:45.340 align:middle line:90% and this one is or natural. 00:44:45.340 --> 00:44:51.210 align:middle line:84% Now we are always comparing apples to apples. 00:44:51.210 --> 00:44:53.740 align:middle line:90% 00:44:53.740 --> 00:44:57.020 align:middle line:84% So this is just a little programming trick. 00:44:57.020 --> 00:45:02.390 align:middle line:84% And in case you really care about the runtime, 00:45:02.390 --> 00:45:05.920 align:middle line:84% I only had to create one extra accidental object. 00:45:05.920 --> 00:45:07.410 align:middle line:90% I think that should work. 00:45:07.410 --> 00:45:09.970 align:middle line:90% 00:45:09.970 --> 00:45:11.000 align:middle line:90% I see some confusing. 00:45:11.000 --> 00:45:12.640 align:middle line:90% Yeah? 00:45:12.640 --> 00:45:15.320 align:middle line:84% AUDIENCE: Why can't we just have the not equal sign? 00:45:15.320 --> 00:45:17.920 align:middle line:84% Like if there's a non-object and it's something that's 00:45:17.920 --> 00:45:19.280 align:middle line:90% not, then will it not compare? 00:45:19.280 --> 00:45:20.280 align:middle line:90% MICHAEL CUTHBERT: Sorry. 00:45:20.280 --> 00:45:23.240 align:middle line:84% Can you say a little bit louder, both for me and others? 00:45:23.240 --> 00:45:24.160 align:middle line:90% AUDIENCE: Yeah. 00:45:24.160 --> 00:45:26.783 align:middle line:84% Once we have the accidental in the key signature-- 00:45:26.783 --> 00:45:27.700 align:middle line:90% MICHAEL CUTHBERT: Yep. 00:45:27.700 --> 00:45:30.160 align:middle line:84% AUDIENCE: --why do we need to have the or natural? 00:45:30.160 --> 00:45:32.470 align:middle line:84% If we're comparing the not equals in the next line, 00:45:32.470 --> 00:45:34.113 align:middle line:90% will that not work with non-- 00:45:34.113 --> 00:45:34.988 align:middle line:90% MICHAEL CUTHBERT: Ah. 00:45:34.988 --> 00:45:39.710 align:middle line:84% Yeah, so one of the reasons might be-- 00:45:39.710 --> 00:45:40.770 align:middle line:90% yeah, that's true. 00:45:40.770 --> 00:45:43.880 align:middle line:90% We probably don't. 00:45:43.880 --> 00:45:45.030 align:middle line:90% Wait, no, no. 00:45:45.030 --> 00:45:47.420 align:middle line:90% Then we get to Matthew's case. 00:45:47.420 --> 00:45:51.830 align:middle line:90% What case does this not cover? 00:45:51.830 --> 00:45:56.187 align:middle line:84% AUDIENCE: I don't see a case that this doesn't cover. 00:45:56.187 --> 00:45:57.230 align:middle line:90% MICHAEL CUTHBERT: Great. 00:45:57.230 --> 00:46:01.280 align:middle line:84% So here's the case that doesn't cover. 00:46:01.280 --> 00:46:03.720 align:middle line:90% ks, again, is our D major thing. 00:46:03.720 --> 00:46:12.545 align:middle line:84% Accidental by step of G is none, right? 00:46:12.545 --> 00:46:17.480 align:middle line:90% n equals note.Note G sharp-- 00:46:17.480 --> 00:46:18.570 align:middle line:90% G sharp four. 00:46:18.570 --> 00:46:19.230 align:middle line:90% Great. 00:46:19.230 --> 00:46:23.960 align:middle line:90% Next n equals note.Note G four. 00:46:23.960 --> 00:46:27.470 align:middle line:84% When I generate this, it's going to get an accidental 00:46:27.470 --> 00:46:29.220 align:middle line:90% because it needs something. 00:46:29.220 --> 00:46:32.640 align:middle line:84% So you're going to end up with next n.pitch.accidental-- 00:46:32.640 --> 00:46:36.350 align:middle line:90% 00:46:36.350 --> 00:46:43.470 align:middle line:84% oops. dot. pitch.accidental equals pitch.accidental natural 00:46:43.470 --> 00:46:46.740 align:middle line:84% because it will actually have a natural sign. 00:46:46.740 --> 00:46:55.430 align:middle line:84% And then this line up here, next n.pitch.accidental. 00:46:55.430 --> 00:47:01.220 align:middle line:90% 00:47:01.220 --> 00:47:07.700 align:middle line:84% Does it equal ks accidental by step G? 00:47:07.700 --> 00:47:10.010 align:middle line:90% Oops, what did I forget? 00:47:10.010 --> 00:47:13.370 align:middle line:84% Didn't I import pitch, or did I-- 00:47:13.370 --> 00:47:15.650 align:middle line:90% I never ran that. 00:47:15.650 --> 00:47:17.990 align:middle line:84% I typed that, but I forgot to run it. 00:47:17.990 --> 00:47:22.310 align:middle line:84% False because this is a natural, an explicit accidental 00:47:22.310 --> 00:47:27.980 align:middle line:90% of natural, and this is a none. 00:47:27.980 --> 00:47:31.000 align:middle line:84% Not a nun like the ones that wear the habits and go around, 00:47:31.000 --> 00:47:33.900 align:middle line:90% but N-O-N-E. 00:47:33.900 --> 00:47:35.817 align:middle line:90% So that's the particular case. 00:47:35.817 --> 00:47:37.650 align:middle line:84% Now, one of the things we never talked about 00:47:37.650 --> 00:47:42.880 align:middle line:90% is how do accidentals compare? 00:47:42.880 --> 00:47:47.700 align:middle line:84% First, we want to make sure does one accidental flat 00:47:47.700 --> 00:47:55.320 align:middle line:90% equals another accidental flat? 00:47:55.320 --> 00:47:57.510 align:middle line:90% Yes, they do. 00:47:57.510 --> 00:48:01.770 align:middle line:84% So that's the reason why I put in these or-- 00:48:01.770 --> 00:48:09.060 align:middle line:84% where'd they go-- or natural, so that I'm always comparing apples 00:48:09.060 --> 00:48:10.350 align:middle line:90% to apples. 00:48:10.350 --> 00:48:11.140 align:middle line:90% Great idea. 00:48:11.140 --> 00:48:12.690 align:middle line:90% Yeah, Jack? 00:48:12.690 --> 00:48:14.870 align:middle line:84% AUDIENCE: I'm confused how the example-- 00:48:14.870 --> 00:48:16.440 align:middle line:90% didn't we want it to be false? 00:48:16.440 --> 00:48:18.780 align:middle line:84% MICHAEL CUTHBERT: Didn't we want it to be false? 00:48:18.780 --> 00:48:21.870 align:middle line:84% AUDIENCE: In the extra thing you just wrote below. 00:48:21.870 --> 00:48:24.300 align:middle line:84% MICHAEL CUTHBERT: In the extra thing? 00:48:24.300 --> 00:48:26.190 align:middle line:90% AUDIENCE: Yeah, that. 00:48:26.190 --> 00:48:27.190 align:middle line:90% MICHAEL CUTHBERT: Right. 00:48:27.190 --> 00:48:33.220 align:middle line:84% We want this one to be false because we're in D major, right? 00:48:33.220 --> 00:48:36.350 align:middle line:90% Accidental on G is none, right? 00:48:36.350 --> 00:48:41.300 align:middle line:84% So this is none, and this is a natural. 00:48:41.300 --> 00:48:43.700 align:middle line:84% But the note is still going to be a G natural. 00:48:43.700 --> 00:48:46.810 align:middle line:84% It's just one of them is represented with no accidental, 00:48:46.810 --> 00:48:50.500 align:middle line:84% and the other is represented with the accidental of natural. 00:48:50.500 --> 00:48:51.040 align:middle line:90% So-- 00:48:51.040 --> 00:48:52.457 align:middle line:84% AUDIENCE: Oh, you're saying having 00:48:52.457 --> 00:48:55.120 align:middle line:90% the natural sign [INAUDIBLE]? 00:48:55.120 --> 00:48:58.100 align:middle line:84% MICHAEL CUTHBERT: So by putting the or natural-- 00:48:58.100 --> 00:49:00.760 align:middle line:90% so I've defined this now. 00:49:00.760 --> 00:49:09.100 align:middle line:84% Acc equals next n's pitch.accidental or natural. 00:49:09.100 --> 00:49:10.670 align:middle line:90% What's acc now? 00:49:10.670 --> 00:49:12.910 align:middle line:90% It's a natural. 00:49:12.910 --> 00:49:20.410 align:middle line:84% ks-- key signature's accidental equals ks accidental by step. 00:49:20.410 --> 00:49:29.150 align:middle line:84% G or natural ks accidental is now also a natural. 00:49:29.150 --> 00:49:34.550 align:middle line:84% Acc equals ks accidental is now true. 00:49:34.550 --> 00:49:40.070 align:middle line:84% The notion that you don't care about the difference between a G 00:49:40.070 --> 00:49:42.200 align:middle line:84% that has no accidental and a G that 00:49:42.200 --> 00:49:44.760 align:middle line:84% has an accidental of natural, right? 00:49:44.760 --> 00:49:49.940 align:middle line:90% In this case, we do not. 00:49:49.940 --> 00:49:54.420 align:middle line:84% So does that start to clarify things? 00:49:54.420 --> 00:49:55.710 align:middle line:90% OK. 00:49:55.710 --> 00:49:57.390 align:middle line:90% Here are some notes. 00:49:57.390 --> 00:50:00.030 align:middle line:84% Let's always go back to actual music. 00:50:00.030 --> 00:50:03.000 align:middle line:84% That's how this-- that's why we're confused. 00:50:03.000 --> 00:50:05.490 align:middle line:84% We haven't looked at some actual music in a little while. 00:50:05.490 --> 00:50:07.170 align:middle line:90% So we're in D major. 00:50:07.170 --> 00:50:08.810 align:middle line:90% Here, we have two sharps. 00:50:08.810 --> 00:50:11.070 align:middle line:84% Let's ignore the time signature for now. 00:50:11.070 --> 00:50:14.213 align:middle line:90% And now we have something like-- 00:50:14.213 --> 00:50:17.200 align:middle line:90% [MARKER SCRIBBLING] 00:50:17.200 --> 00:50:32.200 align:middle line:90% 00:50:32.200 --> 00:50:39.180 align:middle line:84% How many notes here contradict the key signature 00:50:39.180 --> 00:50:41.062 align:middle line:90% if we're looking at this? 00:50:41.062 --> 00:50:41.645 align:middle line:90% AUDIENCE: One. 00:50:41.645 --> 00:50:42.562 align:middle line:90% MICHAEL CUTHBERT: One. 00:50:42.562 --> 00:50:44.470 align:middle line:90% Which one? 00:50:44.470 --> 00:50:46.320 align:middle line:90% That G sharp. 00:50:46.320 --> 00:50:48.640 align:middle line:90% But this has a-- 00:50:48.640 --> 00:50:52.373 align:middle line:84% the accidental on this note is probably none. 00:50:52.373 --> 00:50:55.020 align:middle line:90% 00:50:55.020 --> 00:50:59.755 align:middle line:84% The accidental on this note is almost certainly natural. 00:50:59.755 --> 00:51:02.700 align:middle line:90% 00:51:02.700 --> 00:51:04.860 align:middle line:90% And the key signatures-- 00:51:04.860 --> 00:51:06.480 align:middle line:90% give me another color. 00:51:06.480 --> 00:51:12.520 align:middle line:84% The ks of-- it would go right here-- 00:51:12.520 --> 00:51:20.523 align:middle line:84% the accidental by step for G is none. 00:51:20.523 --> 00:51:23.170 align:middle line:90% 00:51:23.170 --> 00:51:28.390 align:middle line:84% So in this case, these two would compare equal, 00:51:28.390 --> 00:51:29.775 align:middle line:90% but these two would not. 00:51:29.775 --> 00:51:34.480 align:middle line:90% 00:51:34.480 --> 00:51:37.040 align:middle line:84% Not a very good instructor of software engineering today. 00:51:37.040 --> 00:51:37.820 align:middle line:90% Sorry about that. 00:51:37.820 --> 00:51:39.200 align:middle line:90% But I'll try to get better. 00:51:39.200 --> 00:51:41.050 align:middle line:90% Any other questions? 00:51:41.050 --> 00:51:43.160 align:middle line:84% We have about four or five minutes. 00:51:43.160 --> 00:51:47.990 align:middle line:84% I'd love to try to get a little bit further in this, 00:51:47.990 --> 00:51:50.150 align:middle line:90% so I'm going to jump back. 00:51:50.150 --> 00:51:51.980 align:middle line:84% Oh, we have something I never did. 00:51:51.980 --> 00:51:56.050 align:middle line:90% Let's just make it-- 00:51:56.050 --> 00:51:59.590 align:middle line:84% in the interest of time, we'll create a counter 00:51:59.590 --> 00:52:03.550 align:middle line:84% and just keep track of downbeat versus non downbeat. 00:52:03.550 --> 00:52:09.190 align:middle line:84% So what we'll say is if the accidental is not the same 00:52:09.190 --> 00:52:13.400 align:middle line:84% as the key signature accidental, then it's an interesting note. 00:52:13.400 --> 00:52:16.090 align:middle line:84% So then we want to know if it's on the downbeat 00:52:16.090 --> 00:52:17.360 align:middle line:90% or off the downbeat. 00:52:17.360 --> 00:52:19.990 align:middle line:84% What do y'all use to figure out if something was 00:52:19.990 --> 00:52:22.490 align:middle line:90% on the downbeat? 00:52:22.490 --> 00:52:23.190 align:middle line:90% What'd you say? 00:52:23.190 --> 00:52:23.510 align:middle line:90% AUDIENCE: Offset. 00:52:23.510 --> 00:52:24.990 align:middle line:90% MICHAEL CUTHBERT: Offset, right. 00:52:24.990 --> 00:52:28.110 align:middle line:90% So if the note's offset-- 00:52:28.110 --> 00:52:30.843 align:middle line:84% and here is where recurse and flat are going 00:52:30.843 --> 00:52:32.010 align:middle line:90% to be very different, right? 00:52:32.010 --> 00:52:39.240 align:middle line:84% Because recurse's offset is with respect to what? 00:52:39.240 --> 00:52:39.740 align:middle line:90% Vincent? 00:52:39.740 --> 00:52:40.130 align:middle line:90% AUDIENCE: Measure. 00:52:40.130 --> 00:52:41.380 align:middle line:90% MICHAEL CUTHBERT: The measure. 00:52:41.380 --> 00:52:47.150 align:middle line:84% And flatten's offset is with respect to what, Misha? 00:52:47.150 --> 00:52:48.380 align:middle line:90% AUDIENCE: The start of the-- 00:52:48.380 --> 00:52:49.925 align:middle line:84% MICHAEL CUTHBERT: The start of the-- 00:52:49.925 --> 00:52:50.930 align:middle line:90% AUDIENCE: The score. 00:52:50.930 --> 00:52:51.200 align:middle line:90% MICHAEL CUTHBERT: Yep. 00:52:51.200 --> 00:52:52.825 align:middle line:84% Start of the piece, start of the score. 00:52:52.825 --> 00:52:54.480 align:middle line:84% Start of, yep, of whatever it is. 00:52:54.480 --> 00:52:57.420 align:middle line:90% So this one, we want this here. 00:52:57.420 --> 00:53:02.460 align:middle line:84% If the offset is 1.0, we can say accidental positions. 00:53:02.460 --> 00:53:07.700 align:middle line:84% I'm going to do that for, I don't know, downbeat. 00:53:07.700 --> 00:53:13.640 align:middle line:90% Plus equals one else accidental. 00:53:13.640 --> 00:53:16.050 align:middle line:90% Did I somehow not type properly? 00:53:16.050 --> 00:53:17.000 align:middle line:90% There we go. 00:53:17.000 --> 00:53:25.480 align:middle line:84% Dental positions offbeat plus equals one. 00:53:25.480 --> 00:53:30.010 align:middle line:84% Oh, the other thing that we might want to do 00:53:30.010 --> 00:53:33.580 align:middle line:84% is eventually normalize this for how many notes 00:53:33.580 --> 00:53:37.730 align:middle line:84% are on the downbeat versus how many notes are on the offbeat. 00:53:37.730 --> 00:53:39.560 align:middle line:84% You all know about normalization, right? 00:53:39.560 --> 00:53:42.700 align:middle line:84% You want to figure out how much more likely is it. 00:53:42.700 --> 00:53:45.610 align:middle line:84% No, but I can look at the time and see 00:53:45.610 --> 00:53:51.110 align:middle line:84% we're not going to be able to do all that, so yeah. 00:53:51.110 --> 00:53:56.860 align:middle line:84% So let's-- I think this might work. 00:53:56.860 --> 00:54:00.610 align:middle line:84% So let's look at the accidental positions. 00:54:00.610 --> 00:54:03.050 align:middle line:84% And I'm only looking at 3/4 ones. 00:54:03.050 --> 00:54:09.160 align:middle line:84% You might be-- anybody get something like that? 00:54:09.160 --> 00:54:11.263 align:middle line:90% Hannah, you had some results? 00:54:11.263 --> 00:54:13.180 align:middle line:84% AUDIENCE: Well, I did it a little differently, 00:54:13.180 --> 00:54:15.109 align:middle line:84% but I had a slightly different output. 00:54:15.109 --> 00:54:16.026 align:middle line:90% MICHAEL CUTHBERT: Yep. 00:54:16.026 --> 00:54:20.085 align:middle line:84% But It's kind of great when we have two different people doing 00:54:20.085 --> 00:54:20.710 align:middle line:90% different ways. 00:54:20.710 --> 00:54:22.377 align:middle line:84% So you have a slightly different output. 00:54:22.377 --> 00:54:27.280 align:middle line:84% Did you find something like this in your results or-- 00:54:27.280 --> 00:54:28.530 align:middle line:90% AUDIENCE: I took the average-- 00:54:28.530 --> 00:54:29.780 align:middle line:90% MICHAEL CUTHBERT: The average? 00:54:29.780 --> 00:54:32.160 align:middle line:84% AUDIENCE: --of where it appears in the measure, I guess. 00:54:32.160 --> 00:54:36.220 align:middle line:84% So I guess the rest of what I got was closer to the downbeat, 00:54:36.220 --> 00:54:37.137 align:middle line:90% I think. 00:54:37.137 --> 00:54:38.970 align:middle line:84% MICHAEL CUTHBERT: It was a little bit closer 00:54:38.970 --> 00:54:39.640 align:middle line:90% to the downbeat? 00:54:39.640 --> 00:54:40.140 align:middle line:90% Great. 00:54:40.140 --> 00:54:40.973 align:middle line:90% AUDIENCE: I do know. 00:54:40.973 --> 00:54:42.280 align:middle line:90% It's closer to the middle. 00:54:42.280 --> 00:54:43.905 align:middle line:84% MICHAEL CUTHBERT: Closer to the middle? 00:54:43.905 --> 00:54:45.300 align:middle line:90% Fantastic. 00:54:45.300 --> 00:54:49.840 align:middle line:84% I'll give you one little thing, as we do it, 00:54:49.840 --> 00:54:53.252 align:middle line:84% that might be interesting going forward. 00:54:53.252 --> 00:54:55.210 align:middle line:84% We're not going to have time to implement this, 00:54:55.210 --> 00:55:00.640 align:middle line:84% but I will be unlocking this for you right now for n in-- 00:55:00.640 --> 00:55:05.040 align:middle line:84% let's just look at lady note.Note. 00:55:05.040 --> 00:55:09.623 align:middle line:90% Print n and ends beat strength. 00:55:09.623 --> 00:55:12.210 align:middle line:90% 00:55:12.210 --> 00:55:13.960 align:middle line:84% And we'll look at this a little bit. 00:55:13.960 --> 00:55:15.940 align:middle line:90% And this tells you-- 00:55:15.940 --> 00:55:17.360 align:middle line:90% here, do we have the-- 00:55:17.360 --> 00:55:19.640 align:middle line:84% I think somewhere I have the score. 00:55:19.640 --> 00:55:22.190 align:middle line:84% Again, sorry, I wish I could split this here. 00:55:22.190 --> 00:55:24.530 align:middle line:84% So the first D is on the downbeat, 00:55:24.530 --> 00:55:28.610 align:middle line:84% and then we're going to go up to F natural and the C natural. 00:55:28.610 --> 00:55:30.760 align:middle line:90% And when we look at this-- 00:55:30.760 --> 00:55:39.238 align:middle line:84% oops-- later, we can see we have the three pickup notes. 00:55:39.238 --> 00:55:41.030 align:middle line:84% And then we have something on the downbeat. 00:55:41.030 --> 00:55:43.600 align:middle line:84% And then we have a bunch of offbeat things. 00:55:43.600 --> 00:55:44.740 align:middle line:90% Yep. 00:55:44.740 --> 00:55:50.150 align:middle line:84% Beat strength is just a really, really hacky thing 00:55:50.150 --> 00:55:52.960 align:middle line:84% I threw together in probably about the same amount of time 00:55:52.960 --> 00:55:57.130 align:middle line:84% as this thing here that tries to measure 00:55:57.130 --> 00:56:00.223 align:middle line:84% as a number between epsilon and one. 00:56:00.223 --> 00:56:01.640 align:middle line:84% There is no beat strength of zero, 00:56:01.640 --> 00:56:10.000 align:middle line:84% but how strong the position of the note within its measure is. 00:56:10.000 --> 00:56:13.210 align:middle line:84% I cannot believe how many times I've seen it referenced 00:56:13.210 --> 00:56:13.880 align:middle line:90% in papers now. 00:56:13.880 --> 00:56:16.170 align:middle line:84% I've used it in papers and things like that, 00:56:16.170 --> 00:56:21.710 align:middle line:84% and now the world knows that this was a terrible hack that 00:56:21.710 --> 00:56:24.600 align:middle line:84% could probably be done much, much, much better, 00:56:24.600 --> 00:56:27.980 align:middle line:84% as you will realize when you do beat strengths on triplets 00:56:27.980 --> 00:56:30.410 align:middle line:84% and things because you can see I'm doing something 00:56:30.410 --> 00:56:34.230 align:middle line:84% about the log2 of the proportion through the measure, 00:56:34.230 --> 00:56:37.830 align:middle line:84% and therefore triplets won't work as well or anything. 00:56:37.830 --> 00:56:40.740 align:middle line:84% So I think we do something separate for 6/8 so it works. 00:56:40.740 --> 00:56:45.200 align:middle line:84% But this could be used to give you a little bit more 00:56:45.200 --> 00:56:46.940 align:middle line:84% of a sense of the kind of problem 00:56:46.940 --> 00:56:49.440 align:middle line:84% that Sophia Son was talking about in the video 00:56:49.440 --> 00:56:53.390 align:middle line:84% at the beginning of the class, are certain things happen more 00:56:53.390 --> 00:56:56.390 align:middle line:90% on strong places versus others. 00:56:56.390 --> 00:56:57.310 align:middle line:90% OK. 00:56:57.310 --> 00:56:59.790 align:middle line:84% Can you close what you're working on for just a second? 00:56:59.790 --> 00:57:01.373 align:middle line:84% And that's also the cue that I'm going 00:57:01.373 --> 00:57:03.840 align:middle line:84% to let you go just so I can have some eyes and things. 00:57:03.840 --> 00:57:07.470 align:middle line:84% So the next problem set-- you're working on this one, 00:57:07.470 --> 00:57:09.590 align:middle line:84% but just don't pick up for one second-- 00:57:09.590 --> 00:57:14.940 align:middle line:84% is going to be trying to find yourself in interesting corpus. 00:57:14.940 --> 00:57:17.880 align:middle line:84% Whether or not you decide to end working 00:57:17.880 --> 00:57:20.170 align:middle line:84% on a corpus study at the end of the semester or not, 00:57:20.170 --> 00:57:23.370 align:middle line:84% it's probably good to know that we do not want to just continue 00:57:23.370 --> 00:57:25.800 align:middle line:84% working with the Bach Chorales, and the Irish folk songs, 00:57:25.800 --> 00:57:27.675 align:middle line:84% and the Chinese folk songs-- these three data 00:57:27.675 --> 00:57:30.130 align:middle line:84% sets that are everywhere-- for the rest of our lives. 00:57:30.130 --> 00:57:31.540 align:middle line:90% So we want to have that out. 00:57:31.540 --> 00:57:33.400 align:middle line:84% You're going to be asked to do that 00:57:33.400 --> 00:57:35.440 align:middle line:84% and to write up what you find about something, 00:57:35.440 --> 00:57:37.690 align:middle line:84% and you're going to be asked to do it in teams of two. 00:57:37.690 --> 00:57:42.090 align:middle line:84% Your only condition is that you cannot be on a team with 00:57:42.090 --> 00:57:44.650 align:middle line:84% somebody that you did the last group problem set with. 00:57:44.650 --> 00:57:48.400 align:middle line:84% So find somebody who is not in that problem set, and do that. 00:57:48.400 --> 00:57:50.500 align:middle line:84% And I wanted to give you some extra time for that. 00:57:50.500 --> 00:57:51.810 align:middle line:90% Great. 00:57:51.810 --> 00:57:55.020 align:middle line:84% It's a little bit more on the Sofia Sun and my article online 00:57:55.020 --> 00:58:01.445 align:middle line:84% that you can watch today in the Canvas or later, and thank you. 00:58:01.445 --> 00:58:11.000 align:middle line:90%